Cell Reporting and Object Management Environment

CROME OVERVIEW AND BUSINESS CASE (12/4/2006)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Quantum

TRANSITIONERS OF TECHNOLOGY

CROME
OVERVIEW AND BUSINESS CASE
Copyright © 1997-2006
Quantum Systems Integrators Inc.  All rights reserved.


TABLE OF CONTENTS

 

Introduction.. 1

About CROME. 1

The CROME Client.. 1

The CROME Server.. 2

About Quantum... 2

Cost Benefit Analysis.. 3

Real World Cost Benefit Example. 4

Best in the World Performance. 4

On Demand Reporting.. 5

Control and Consistency.. 6

Empowerment.. 6

Resource Analysis and Capacity Planning.. 7

Wireline planning.. 7

Wireless RF planning.. 8

Busy Periods. 10

Summary.. 11

CROME open Reporting API 11

Built in CROME client export functions. 11

CROME client.. 11

Built in WEB/HTML/GIF export.. 13

Built in CROME Server Functions. 16

OSI NetExpert.. 16

Netcool/OMNIbus®.. 16

The formulas/thresholds used in the variance reports. 17

JAVA Servlets. 18

Geographic Mapping of Sites with CROME Variance Data and Foreign Data.. 20

Address Lookup. 23

Site Finder.. 23

HDML and WAP. 25

Auto Reports and Batch Reporting.. 27

Summary.. 27

Report and Formula Creation Flexibility.. 28

Ad-Hoc Formula Creation: 28

Ad-Hoc Report Creation: 28

Empowerment and Standardization.. 28

Report Generation Flexibility.. 28

The physical dimension: 28

The time dimension.. 28

Any definable range (from a half hour to a decade) 29

With or without weekends or holidays. 29

Special defined service periods or busy hours, for instantaneous reporting.. 29

"Busy Hour on the Fly" from any report, for any formula, for any period.. 29

Customized thresholds, sorts and limits. 29

Drill Up, Same, Down.. 30

Generic, System Independent and Extensible. 33

iDEN OMC system support.. 33

iDEN DAP system support.. 34

iDEN MDG system support.. 35

Nortel DMS family system support.. 36

Nortel GSM OMCR family system support.. 37

Nortel DMS-MTX CMDA family system support.. 38

Other TBD System Types. 39

Summary.. 39

Corporate/National Reporting.. 40

Geographical Organizations. 40

Automatic Maintenance of Geographical Groupings. 41

CROME Designed for National Scalability.. 42

Advanced Analysis Features.. 43

Migration Tracking.. 43

Grouping.. 43

GIS reporting.. 45

Neighbor Cells and Airgen Worksheets. 46

Neighbor Cells. 47

Airgen Worksheets. 48

GIS reporting by arbitrary area (GIS Grouping) 51

Extendable to non-PM domains (e.g. Drive Test Analysis) 53

Full interpretive Scripting via Plug-In Support.. 58

Erlang and GOS support.. 59

Erlang B.. 59

Erlang C.. 59

Heterogeneous System Support.. 60

Scalable. 60

Fault Tolerant for 100% uptime. 61

Zero Client Administration.. 61

Installation of the CROME application.. 61

NetworkUpdate. 63

Built in Help and Support Email Gateway.. 63

Total System Flexibility.. 64

Testimonials for QSI Support by QSI Customers.. 67

Customer supplied "Cost Benefit" results.. 71

CROME Performance and Scaling (iDEN OMC) - Benchmark run on June 12, 2002. 72

CROME Performance and Scaling (Nortel MSC) - Benchmark run on June 12, 2002. 73

 


Introduction

The purpose of this document, "CROME Overview and Business Case", is intended to provide an overview and a business case for the acquisition of the CROME system by your organization via brief introductions and descriptions of the various capabilities of the CROME system.

This overview document does not provide a complete evaluation.  Quantum also recommends that your organization review the "CROME Client Users Guide" which describes with the initial steps of installing CROME as a native application and invoking CROME.  The "CROME Client Users Guide" covers each of the major features of the client GUI from running reports, editing/creating reports, editing/creating formulas, editing/creating groups, and includes other sections that assist in the process of creating extremely complex and powerful reports.

About CROME

CROME (Cell Reporting and Object Management Environment) is a high performance fifth generation client/server distributed DSS system custom crafted to meet the needs of wireless telecommunications providers.

CROME is a performance management (PM) tool that rapidly solves the needs of real-time, historical, and long-term trend analysis of statistical performance of a national wireless network infrastructure.

CROME can measure trends and identify problems down to the level of an individual radio or channel in a wireless network, thus it is invaluable for assisting RF engineers and system planners. However, the true power of CROME is seen in its ability to quickly and efficiently mine and present low level data utilized by engineers into comprehensive, concise national level reports used by national staff, including marketing, sales and top corporate executives.

CROME optimizes all data sources for high speed secure access across an intranet or even the Internet and eliminates both the need for, and the delay from, the traditional data analyst functions.  The CROME system does this by providing easy-to-use exploration and reporting tools to every individual within the organization. Traditional wireless vendors are making far reaching decisions based on intuition because managers and executives have to act fast and cannot get timely information. CROME solves this problem by allowing the decision maker to quickly access information directly in any manner he or she sees fit.

End users (via the CROME client) or automated gateways (via the CROME server) can export data and reports into many formats such as traditional office suites (e.g., Microsoft Excel or Open Office) for distribution and presentation.  In addition CROME supports standard HTML formats for basic web access and interfaces to third party applications such as fault management systems (e.g., NetBoss, HP OVO, Micromuse -or- NetExpert).

The CROME Client

The CROME client is a 100% pure JAVA application/applet for managing the reporting needs of wireless service providers.

CROME reporting can be run from inside a browser (such as Netscape or Internet Explorer) or can be run as a stand-alone application.  When run as an application, there are no browser security restrictions, and thus more features are available  (such as local printing, local file access to save reports and formulas, etc.). Every authorized user in your organization is enabled to run reports when they need them, across time periods they require, and with their own custom groupings or filters.

Operationally, CROME provides the means to quickly produce reports that have been defined locally by the user or stored as “system-wide” reports by the CROME administrator.

The report editing capabilities are extremely flexible, with a point-and-click interface to format the report and to select the appropriate raw statistical data (i.e. pegs) or formulas.  Formulas themselves are defined with the same point-and-click interface, and can contain a variety of pegs, mathematical functions, constants, and even formulas-within-formulas.

Even after a report and its formulas are defined, the same report can display information in hundreds of different ways, by dynamically selecting choices when running the report.  Such choices include time granularity (hours, days, weeks, months years), summary options (daily, bouncing-busy-hour, etc.), viewing the data by its network element, and specifying the desired physical network level.

CROME also provides the ability to run any report within an arbitrary grouping of user-defined network elements, and a flexible editor to create the groups (which allows direct selection of known network elements and regular-expression grouping).

The CROME Server

The CROME server is a distributed and highly scalable data warehouse / decision support system that provides near instant enterprise access to key statistical performance data.  Together one or more CROME servers and at least one CROME client or API gateway comprise a CROME system.

A set of CROME servers can also be configured to mirror each other such that they form a fault tolerant "ring" for high availability and non-stop operation.  In such a configuration, a failure of an individual server is completely transparent to any CROME client.

The CROME design offers orders of magnitude performance over traditional RDBMS solutions via proprietary space and time optimizations - yet delivers its data in a multitude of useful formats to any desktop or web-top within the enterprise.  CROME servers maintain data for time spans ranging from six months to decades (depending upon available storage) providing a framework for powerful historical and trend analysis.

The CROME open Gateway API is an open solution for data import, able to acquire data from both equipment and legacy systems including serial, TCP/IP, X.25, direct RDBMS, SNMP and CMIP/Q3 interfaces.  For more information, refer to the document "CROME open Gateway API specification".

The CROME open Reporting API is an open solution for data export allowing easy distribution of data to third party products.  For more information, refer to the document "CROME open Reporting API specification".  Some actual working integration examples using this API are shown in this document.

About Quantum

Quantum is a Systems Integrator and Software Developer for mission critical applications: QSI personnel provide our clients with in depth experience in the development and integration of custom real-time, mission-critical systems for telecommunications, petrochemical and aerospace industries.

Quantum is a leader in applying technology: Using our internally developed APIs and products we are able to offer clients significant technology advantages and cost savings. Developed using object oriented design concepts, the QSI library has been providing clients with:

·         Significantly reduced development time

·         Minimal maintenance for Zero Administration Costs (ZAC)

·         Multi-Platform Software Design and Porting for true portability

·         Client/server computing

·         Internet development CGI & PERL scripts, JAVA, SERVLET and proprietary solutions

·         Real-time "Mission Critical" application development

·         DBMS, Data Warehousing, and data mining

·         Graphical and Character user interfaces

·         Interfaces, i.e. middle-ware, to both relational and object database systems

·         Network Management via CMIP, SNMPv1, and SNMPv2

·         Extensive JAVA application expertise including DBMS, VRML and DSS systems

·         Web hosting, and advanced Web site development via Apache Web server products

·         Hardware and Software Sales (DBMS/Workstations/Peripherals/Software)

Quantum has leveraged over 30 man years of experience in the creation of both wireless and wireline performance reporting tools into the development of the CROME product line to meet the unique and demanding reporting requirements of telecommunications operators.

For further information on Quantum, CROME, or to establish a full scale Demonstration/Validation deployment prior to purchase, contact the Wireless Products Division at Quantum Systems Integrators, Inc. at (714) 428 1133 -or- E-mail wp@quantumsi.com .

Cost Benefit Analysis

As well as providing an overview of CROME capabilities, this document is also meant to illustrate how CROME can create significant cost savings for a wireless operator. Many of the discussions and examples throughout this document are specifically focused on how the planning and maintenance of a national network infrastructure can be improved.

The attributes of  CROME that saves your organization time and money are as follows:

·         Performance -- reduced time-to-solution:  answers in seconds NOT hours.  For example an ad-hoc "Minutes of Use" report at the site level across the state of Utah typically completes in 7/10th of a second.

·         Data accuracy -- the elimination of time consuming error prone tasks.

·         Flexibility -- the ability to be extended, adapted, and integrated with other systems.

·         Access -- the ability to reach the data and gain meaningful insights from anywhere in the enterprise.

·         Needs -- support specific needs such as multiple "customer" busy hours crafted to specific service types.

·         Versatile -- Supports "non-standard busy hours on the fly" maps spreadsheet data directly to GIS views.

·         Scalable -- the ability to grow to larger and larger deployments.

·         Commercial – commercial, off-the-shelf products are more reliable, cheaper to maintain, and in many cases have new features introduced at no cost due to other customer requirements.

·         Proven -- One low end CROME server has operated in a production capability at Nextel for three years, easily handling about 16% of all Nextel US infrastructure in a true distributed demo.

·         Availability -- CROME deployed in multiple locations world wide in1998, the product exists today and can meet your needs "as is". It is not "vapor ware" nor is it a an equipment vendor's dream

Typical uses of CROME that provide positive measurable return on investment:

·         CROME can be used to optimize major equipment resources, at the proper busy hour, to minimize procurement and installation (for example, help accurately determine how many Base Radios or DS0s are really needed to meet the desired grade of service to the customer - or - planned build-out). This lowers and delays capital expenditures.  Note the customer can define multiple business busy hours or any formula can become a busy hour on the fly to cover special cases.

·         The speed or time-to-solution of CROME (get answers in seconds instead of hours) drastically reduces the engineering hours necessary to perform critical analytical tasks.  This lowers labor costs and enables better planning, more detailed forecasting and "what-if" analysis.

·         Standardization: the CROME system both promotes and enforces standards, increases organizational memory and efficiency.  In many cases reports can encapsulate you companies best "expert knowledge" and thus disseminate best practices throughout the organization.

·         CROME's performance also allows near real-time reporting and proactive correction of out of variance situations.  This improves customer satisfaction.

·         The CROME open Reporting API allows ease of integration into both analyst’s PC desktop applications as well as server-side data and object management systems, minimizing the time users must spend with other local and enterprise software.

·         Data sharing supports a "single service" concept of data distribution in the enterprise eliminating hard to maintain redundant application islands.  CROME can read information from other central data service points as well as perform its primary role as the performance management (PM) service.

·         Integration with legacy systems (in both data input and data output) minimizes upgrade costs in other subsystems and enhances the lifetime and usability of these systems.

·         CROME opens up possibilities to proactively and automatically tune the system to meet specific one time events (e.g., the Olympics) or marketing promotions.

Other:

·         Elimination of BIG iron costs, the CROME solution does not need expensive multi-million dollar platforms like the high end Sun's E25K data center platform to operate or scale to larger and larger deployments.

·         Elimination of hidden costs like RDBMS licenses, which grow quickly at high user counts, and/or extremely expensive CPU based licensing.  The speed of CROME minimizes end user connect time to the RDBMS and thus lowers the cost of ancillary third party software.

·         Very low management, upgrade and support costs. Quantum manages local US and overseas CROME servers from Asia to the Middle East in a completely remote fashion.  All client software is self-updating with built in help and error reporting.

·         Quantum has an unmatched client/vendor relationship.  Simply ask any direct customer of Quantum services.

Real World Cost Benefit Example

In a small five (5) iDEN OMC deployment CROME was used to replace a multi-million legacy performance management system developed by the customer.  After eighteen months of use the customer documented tangible and intangible cost savings as follows:

·         Engineering hours associated with report generation dropped from 70 hours per week to 4 hours per week, estimated yearly savings $210K

·         Near real-time variance reporting allowed proactive correction of problems in hours instead of days, thus improving quality of service to wireless customers.  Allows the network to be operated efficiently with fewer resources at the target grade of service levels.

·         Improved data accuracy and completeness. Data holes are flagged and late and/or missing data is automatically  recollected.  Better and more complete data allows for better decisions.

·         Ability to create and run special needs "ad-hoc" reports in seconds to validate system upgrades and assist in fault isolation.  Flexibility allows for more specialized reporting and better tend analysis and resource optimization.

·         Built in Erlang analysis functions drastically improved resource allocation procedures from simple request counts to true grade of service metrics.  Eliminated guess work and tendency to overbuild infrastructure thus lowering capital expenditures.

Best in the World Performance

CROME is not a simple batch processing system used to generate reports from a flat RDBMS. CROME is a complex, complete data management and data mining environment offering an unprecedented level of performance over traditional batch processing and direct SQL based queries. CROME offers outstanding performance whether doing a quarterly report across a multi-switch national area or reporting on all available data across thousands of sectors for any given day. Simply put, CROME does what our competitors can only claim.  When using CROME as your reporting tool, full custom ad-hoc market and region level reports are complete in a few seconds and national reports across scores of switches in a few minutes.

CROME reports are always fully user definable and are generated in a matter of seconds - most competing systems require lengthy batch processing runs taking several hours.  CROME is an ideal system to drastically improve time-to-solution for performance management reporting.

Examples of CROME's performance during full ad-hoc reporting follow:

·         CROME can analyze ten (8) Nortel MSC (with 418 trunk groups) during a specific bouncing busy hour in less than one half of a second using built-in Erlang-B analysis and report on all trunks failing to meet a specific grade of service (actual full ad-hoc report took 0.47 seconds on a LAN).

·         CROME can supply minutes of use reports (interconnect / dispatch / total) across a fully loaded Motorola iDEN OMCs (with 682 sectors) during a specific bouncing busy hour in less than one (1) seconds (actual full ad-hoc report took 0.64 seconds on a LAN).

·         CROME is highly scalable can supply minutes of use reports (interconnect / dispatch / total) across five (5) fully loaded Motorola iDEN OMCs (with 4,609 sectors) during a specific bouncing busy hour in less than five (6) seconds (actual full ad-hoc report took 5.92 seconds on a LAN).

On Demand Reporting

The CROME database (or a network of distributed CROME databases) is the only source of data for the CROME client.  There is no intermediate files or redundant data.  Every report is run on demand - in a full ad-hoc fashion - from the most current and accurate data available.

Traditional solutions typically batch-process data and create intermediate files and or tables for later analysis stages.  These systems lead to loosely coupled and poorly managed flows of data.  Traditional batch systems decouple the final step or steps of processing from the original data source, therefore the end-user (or final report) remains unaware of changes due to either corrections of late data.

Similarly, CROME reports selected by an end user utilize "late binding" of user preferences and filters which allows a single generic CROME report to replace up to fifty (50) hard coded scripts or reporting procedures.

Users of the CROME client can select the "Physical Layer" or level in a tree based hierarchy and also select a specific "Summary by" to bind to the generic CROME report.

Another typical CROME filter “View By” controls the period of either by element, group, or the granularity in the time dimension and determines the “x-axis” on the graph presentations to the generic CROME report.

       

Note there are many more filters in CROME which are beyond the scope of this introduction. For more information on the CROME client software, please refer to the CROME Client Users Guide.

Control and Consistency

CROME provides unsurpassed version control of your company’s critical reporting, yet still allows ad-hoc analysis by any engineer in the organization. Thanks to CROME's built-in version stamping and source stamping for both formulas and reports and compatibility tracking system, you won’t find yourself presenting the wrong data to your management due to an outdated formula.

Standard reports can be defined and mandated.  All reports include version stamping to allow the organization to unambiguously identify what report and what formula set was used to ensure that, where appropriate, approved corporate standards are followed.  However end users are still free to experiment, their reporting outputs are simply stamped and versioned as a "local" non-standard or non-approved result.

Since all released versions are maintained on the CROME server, comparisons between various revisions or "apples to apples" analysis can be done to validate and/or prove superior formula/reporting sets.

The design goal of CROME promotes organizational uniformity and standards but does not hamper the engineer or manager in performing special ad-hoc analysis quickly, in localized contexts, to respond quickly to local emergencies or perform local "what if" analysis.

Empowerment

Because CROME client software is 100% JAVA, PC users now have the power of a true data mining application with the performance and flexibility previously only offered by expensive workstations. Installation onto any Win95/98/2000/NT/XP PC platform is done via a simple click from a Web page. No longer is there an analyst between you and your data. CROME frees the analyst from his typical role, a single interface between reporting tools and a multitude of report generation requests, to his proper role of validating formulas and designing reports.

CROME allows every authorized user in your organization to run reports when they need them.  This level of empowerment allows non-engineering departments such as finance, sales, and marketing to get important data instantly.  For example there may be a financial need to track radio counts in the field that might take weeks (and hours of phone calls) to regional markets - this information is instantly served by CROME. We have found that CROME customers constantly find new and novel uses for enterprise data.

Because CROME delivers answers to questions swiftly, your company makes the best use of its statistical and performance data to proactively improve its current operations, strategically plan and budget future resource needs, and avoid costly mistakes.

Resource Analysis and Capacity Planning

CROME includes precise high speed mathematical features to quickly generate "one button" capacity planning reports.  One typical use of CROME reports is in resource and capacity planning.  Built-in Erlang B, Erlang C and grade of service (gos) functions instantly and accurately identify resource bottlenecks in wireless and wireline networks.

CROME also allows "what if" analysis.  For example a report/formula set can be modified to adjust channel reservation parameters or "drop" one radio from all sectors in an RF network to view the possible impact of provisioning changes, or perhaps over provisioning, of RF resources.

Wireline planning

CROME can analyze dozens of Nortel DMS switches at various busy hours (either customer defined or via the advanced "busy hour on the fly" analysis mode) in a matter of seconds and generate a report showing only those trunk groups (or routes) that are under provisioned.   This type of report can tell the end user exactly how many DS0's must be added to each under-provisioned trunk group to meet a user specified grade of service target.

The above report (Real data from 5/3/02 for 1/8th of all Nextel Communications, Inc. subscribers) took under one (0.83) seconds to run across 1141 trunk groups in twelve fully burdened Nortel DMS switches. The resulting CROME report identified nine routes which require additional DS0 circuits to meet a specified grade of service level. [1]

The above report shows data "if and only if" the number of working circuits (WCCT) is less than the justifiable number of DS0s (WCCT_TRU_JSTD) required to meet the customers 1% grade of service target (WCCT_CAR_TRU_GOS) for voice traffic (i.e. TRU from the OM group TRK).

Of course all the functions of the CROME client are still available, such as sorting, limiting and "on the fly" creation of new experimental thresholds.  The "Report Item" menu bar allows a variety of options to be applied to or against the current report or "Report Item".  The standard "Report Item" Menu Bar is shown below:

There is also a matrix Report Item Menu Bar.  If, when running the report from the main screen, you choose “Plot Elements” or “Plot Groups”, the Menu Bar for the Report Item looks a little different (note, in this mode "busy hours on the fly" can be enabled where every formula in a given report can be instantly selected for a busy hour view on that formula):

For more information on the CROME client software "on the fly" functions, please refer to the CROME Client Users Guide.

The same data is simple to export into other tools such as Microsoft Excel or Sun Microsystems' Star Office.  Below is the same data exported via the "File/Export/Result Set" menu operation.

Note, this report instantly under one (0.92) second while performing extensive Erlang-B analysis, an intensive mathematical summation of factorial results, on every trunk group and found the minimal number of DS0's required to meet a defined grade of service metric during a voice traffic busy hour.

It is also important to realize that no manual Erlang or probability table lookups by your RF engineers were needed all the required mathematical capabilities for capacity planning are built into the CROME client application.

Wireless RF planning

CROME can also dynamically analyze multi-function RF resources to determine and report the optimum radio counts per sector in a wireless network and allow fast "what if" analysis when varying pre-allocated and dynamic resource constraints for various RF allocations.

The iDEN environment where different service types (i.e., dispatch, interconnect, and packet data) co-exist in the same RF environment can benefit drastically from CROME analysis.  Each service type has its own unique busy hour due to varying customer peak usage.  In addition, when calculating the grade of service, each service type requires specific sets of formulas to correctly analyze that service type.

CROME allows every service type to be correctly analyzed with its own specific formulas at its own unique busy hour.  For example, a result of a dispatch grade of service analysis is shown below that takes into account the impact from the reserved channels for interconnect (i.e. the provisioning item i3tchminnum).

The above report (Real data from 2/18/02 for 1/8th of all Nextel Communications, Inc. USA subscribers) took under thirty (29.6) seconds to run across all sectors in five (5) iDEN OMCs (with 4,464 sectors).  Once this CROME report was finished it was then was exported to Microsoft Excel with a simple menu selection.  The applied threshold and sort shows the sectors with the worst grade of service at a dispatch busy period of 30 minutes (a 30 minute busy period has higher peaks than a busy hour) sorted by the formula "gos_t0".  This analysis is used for dispatch and packet data capacity planning in an iDEN environment (a different report is required to handle normal interconnect calling grade of service levels).

Note, this report performed extensive Erlang-C analysis on every sector and found the actual measured grade of service (“gos”) at several different call delays while taking into account reserved channels and dynamic channel allocation between service types in an iDEN environment.  It is also important to realize that once again no Erlang or probability table lookups were required - all the required mathematical capabilities are built into the CROME client application.

Busy Periods

One of the most important features of CROME is that your organization can define the actual service windows (i.e., fixed windows of time) and the formulas used to calculate busy periods, typically called a bouncing busy hour.  In addition, CROME allows any number of important service windows or busy periods to be defined.

In the iDEN environment this ability is critical for correctly assessing and analyzing all service types.  For example the dispatch busy period is NOT the same as the interconnect busy period and thus doing any resource planning on the wrong busy period can lead to expensive and costly mistakes.

A quick illustration via an example of running a capacity planning report on (1) the WRONG busy hour and then on (2) the RIGHT busy hour, will now be discussed to highlight the need for multiple busy hours.

Below is a gos report built for dispatch analysis BUT incorrectly run on an interconnect busy hour (it is sorted by gos_t0 and limited to the top 10 items – the below report took 7.2 seconds across 1,707 sectors):

Below is the same gos report built for dispatch analysis BUT correctly run on an dispatch busy hour (it is sorted by gos_t0 and limited to the top 10 items  – the below report took 7.2 seconds across 1,707 sectors):):

It should be apparent that the results are off between the two reports by about 300%.  It is critical to have the proper tools to analyze your data in the proper periods.  Competitive solutions that only provide one busy period will result in mistakes with a very high cost in either capital equipment and/or customer satisfaction.  The CROME system allows multiple busy periods to be defined to meet all your operational needs.

Summary

CROME provides instant reporting, multiple busy period support, and eliminates tedious table lookups for functions like Erlang B and Erlang C.  These key mathematical features coupled with high performance frees the RF engineers in your organization from time consuming tedious and error prone manual work.  In addition, more exotic tuning and planning equations can be quickly crafted to meet explicit environmental or strategic constraints.

In addition, CROME can be configured to also access and utilize key provisioning constants such as i3tchminnum and dchprefnum to facilitate exotic calculations when RF resources are shared between service types.

In CROME many different busy periods can be defined to suite your organizational needs, below is a set of both busy hours and busy half hours that were created and applied to both a Motorola iDEN infrastructure and a Nortel OMCR infrastructure.

  

Busy periods are simply formulas defined by the CROME administrator to create a DSS Cube for ultra high speed analysis.  The CROME also allows foreign busy periods to be pushed onto a different system type, for example a peak MSC utilization can :

CROME open Reporting API

The CROME open Reporting API is an open solution for data export allowing easy distribution of data to third party products.  For more information customers should refer to the document "CROME open Reporting API specification".

This section will highlight the use of the "CROME open Reporting API" and currently existing integration with a variety of sub-systems and data distribution types.

Built in CROME client export functions

CROME's basic User Interface (UI) can directly export data (with or without formulas applied) to flat files or to spreadsheets like Microsoft Excel, Microsoft Works, and Sun Microsystems Star Office.  This allows easy publishing and dissemination of reports across the enterprise using the familiar PC-based products your organizational staff works with every day.

CROME client

Many of the examples in this document show Excel style exports, which are generated from the CROME client via a simple menu selection, at which time the Excel application is invoked and data is automatically populated into a worksheet.  Quantum also provided macros to graph CROME data within Excel via a hot key.

Because the output of CROME reports are sent to the default registered viewers in a Win95/98/2000/NT/XP environment, once CROME is installed you can start mining data to your specifications and seamlessly transfer the results into office productivity tools such as Microsoft Excel or PowerPoint via bridging capability built into the CROME client.

Note, integration with UNIX based office suites such as Star Office and various Linux/KDE/Gnome implementation have all been successfully supported by the CROME GUI client in production CROME deployments by Quantum.

Some examples of MS-Office integration

Of course the data can be manipulated in whatever fashion as true numeric values once exported.

For more detailed and expanded information on CROME client capabilities please refer to the CROME Client Users Guide.

Built in WEB/HTML/GIF export

CROME's open Report API directly supports automatic generation of graphs, spreadsheets, flat files and GIF images for automatic publishing to a standard web server.

Typically, we recommend the use of the CROME client, however there are instances where organizations want to have a few standard reports with fixed filters to be automatically and permanently published on a web server.  Another possibility or reason to use CROME's web export interface is to publish quality of service or uptime reports for a specific customer (or even the general public) on a non-secure www server exposed and available to the Internet.

Shown below is CROME Web Reports, and feature of CROME that provides HTML-based CROME generated report/GIF sets that can be placed on any web server for instant access (without the Java based CROME client) from any browser.

The Web Reports main selection screen:

Example Web Report Graph View:

Example Web Report Result Set View:


Built in CROME Server Functions

The CROME open Report API directly supports in-house analysis tools to create outputs such as “top N” sorted variance reports, and provides a performance management hooks to other third party products.

The CROME system has been providing real-time PM variance reporting for directly OSI's NetExpert™ (as shown below). The integration of NetExpert was performed in a matter of hours by Quantum System's staff in Costa Mesa.  Products such as Micromuse's Netcool® (or any other TBD Fault Management System) can also be quickly be supported using the CROME open Reporting API.

Note the same real-time out of variance information is also available now on a live, self-updating web page and via “HDML”, which enables any packet data HDML compliant iDEN phone to be a CROME data browser (See HDML and WAP section below).

OSI NetExpert

Below is a typical NetExpert display used with reports and thresholds made to the exact requirements and specifications of a Quantum customer. 

Netcool/OMNIbus®

Below are a few a typical displays from Netcool®, another fault management engine, also fed with the CROME reports and thresholds. This is a prime example of PM to FM integration, via generic export gateways of CROME.

Please note the power of the CROME open Report API.  This initial Netcool integration task was accomplished by a non-developer from Micromuse in a few hours via simple E-mail instructions (i.e. no manual was required).  Note this is not yet a full implementation but a quick validation to prove that data can in fact be easily exported from CROME into yet another system.

The formulas/thresholds used in the variance reports

The CROME Real Time data is generally 35-90 minutes behind the current time, since the most network elements bundle performance management data into units of time (e.g. statistical aggregation of pegs or counters).

Any ad-hoc report created in CROME can be configured and placed in a set of as a real-time variance report (to be run every ½ hour or every fifteen minutes to match the emission granularity of data).

As another example of flexibility, CROME can even be configured to produce variance reports like "No Interconnect or Dispatch Traffic" report ONLY between the hours service hours like of 7am and 7pm, because too many sectors report silence during the off peak late night hours. In addition, new sites are automatically filtered out during their first day of operation. Example variances that are tracked in CROME every ½ hour by one of our iDEN customer’s is shown below:

Issues to address in future releases:

1.(Filtering) Predefined sites are not currently masked from the report as there is no current automated centralized repository for information that specifies that sites are built but not active. This capability will shortly be added to the TOAD system at which time CROME will also utilize the centralized exclusion list.

2.(Filtering) Test bed sites are not currently masked (see above).

This data is actually created from a set of standard CROME reports written to find variance situations.  The reports are automatically run every half hour (or alternatively smaller granularities) and streamed into a Quantum Multicast Buffer (or MCB) where it can be accessed by multiple applications (including text and GIS views from QuantumSI tools and third party fault tools).

For example, the same data feed is accessed by TCP/IP by both NetExpert™ and Netcool® fault analysis engines which utilize simple rule sets or probes to respectively analyze the data and generate alerts.

JAVA Servlets

CROME's open Report API directly supports Java Servlets.  Below is a Java-based Servlet hosted on an Apache web server (which serves the same data as seen in the NetExpert™ alert display shown previously).

Quantum constructed a self-updating web page (i.e. it automatically refreshes every 30 minutes) utilizing Java Servlet technology to show variance data information in HTML format.  An example web page is accessible at http://newt.quantumsi.com/jbin/CromeVariance

This web page displays the same data that is streamed into third party FM systems such as NetExpert and Netcool.

Note that each line has a red-green bar showing the number times this Sector was out of variances over the last 48 half-hour reporting periods.  In the above picture the “hand-cursor” mouse is positioned above the 7/48 image for 6255-DirectorsGuild, Sect 3 1 6 0 1 309 85.  This means that this sector was out of variance in 7 of the last 48 half-hour reporting periods.  If you click on this image, you can get a detail page, as shown on the next page:


This page shows the result of clicking on the previous page – all of the variance information for this Sector over the most recent 24 hours.


Geographic Mapping of Sites with CROME Variance Data and Foreign Data

-          This GIS display is actually a highly efficient multi-user big screen display and is used in network operations control centers across the world as well as individual end-user desktops.

The same data shown on the Variance web pages is also shown Geographically in CROME’s Variance Mapping Applet, which also runs directly off of a web page:

This Applet shows all of the EBTS Sites in the area of interest.  The map can be panned and zoomed to find particular sites, and allows the user to light up sites are out-of-variance based on the configured variance formulas (it can combine all the variances on the same map, or even rotate between the different variances at a user-chosen speed).  Like the other variance pages, this is live, current-as-possible real-time variance reporting.

As you move the mouse around the map, the blue text above changes to show the site name, the number of sectors in that site, the number of base radios (BRs) and the number of working BRs, the lat/lon position, and the OMC. 



The applet automatically updates its variance data every half hour and the geographic position of each site every day.

If you click on a site on the map, you will see the sectors and carriers that violated the selected variances.  By clicking on a particular Sector or Carrier in the list, your web browser will pop up the appropriate CROME Variance Detail page, like the one shown on the previous page of this document.  In addition customers how can supply satellite or aerial photography can also integrate addition views.

Note that with this Mapping display you can filter on/off the different OMCs, the different configured variances, whether or not the map shows all sites or just those out of variance, and whether or not to show just sites known to be “On Air” or all sites.

Summary of Features:

-          View the current location of one of our customer sites in Southern California against a geographical map background, the system support and even select between multiple base map styles from various map sources and is capable of importing other custom formats.  Some examples follow:

  
 

-          Pan and Zoom
       

-          View/Hide Sites associated with any given OMC

-          Ability to see, based on the icon and color of the site, whether it has one, two or three sectors.

        

-          Ability to see, based on the color of the site icon, whether or not it is currently out of variance (as near real-time as is available -- usually between 60 and 90 minutes old)

-          Ability to select from one or more configured variances.

-          View/Hide Sites based on Out of Variance or Operating Normally

-          Ability to view all variances at once or automatically "rotate" between them (rotation is typically used to drive a single large Network Operations Display).

-          View/Hide Sites based whether or not the site is On Air or Off Air (this per site indicator is imported from a foreign data source via the generic CROME open API)

-          By simply moving the mouse over a site, you can see the site name, the number of sectors, the number of BRs (configured), the number of Working BRs (i.e. having forward transmit power), the lat/lon position, the OMC, and a textual description of the violated variances, if any.

-          As mentioned previously integration can extend to custom data provided by the wireless operator such as aerial photography.



-          All features of the map display are also available via simple right-mouse-click pop-up  menus:

      

 

-          By simply clicking on a site in the geographic display, all the sectors and/or carriers that are out of variance associated with that site will be loaded into the "out of variance list box".

-          By simply double clicking the sector and/or carrier in the "out of variance list box", you can view the standard CROME Variance 24-hour detail page for that variance item.  This will link and invoke another Web page (refer to: Java Servlets/CromeVariance).  This WEB based linking is setup via the CROME open API.

 

Address Lookup

 

-          CROME Map displays include an Address Lookup feature to help pinpoint geographic locations.  The user can type in any address and see a “house” plotted at the location.  The feature also shows distances from the address to any clicked spot on the map:

 

Site Finder

-          Similar to the Address Lookup Feature, CROME provides a “Site Finder” features to locate an EBTS by name.  A circle is plotted on the map at the location of the site. Like the Address Lookup feature, the distance is shown from the selected site to any mouse-clicked location.

 

 

-          The following section shows yet another way to consume CROME data using an iDEN phone, all utilizing the same open data source and all relying on the same base CROME reports.

-          More examples of map displays follow (e.g. high quality aerial photos)

Top level view above, and 16x zoom below:

 


HDML and WAP

CROME's open Report API directly supports the new HDML and WAP standards. Applications can easily be written to concisely and efficiently disseminate data to current generation wireless phones and/or pagers.

Below is an actual fully functional HDML application suite created by Quantum to highlight both CROME and iDEN phone packet data features. This software suite also includes the same CROME variance reports (previously shown and discussed) but displays the results in the small limited display window of a wireless phone.

A simple "two state" example is shown below. CROME's daily processing completion time stamp (also other metrics) can be quickly accessed via an iDEN phone via various server menu items such as the "Server Status" selection.

 

The only user action required in the above example is to select [OK] for menu choice "3" or "Server Status" and to select [OK] to return back to the main CROME HDML wireless phone menu.

The important point is any data such as key statistical performance data and even near real-time variance data can be viewed from anywhere at anytime via CROME's Open Report API, which allows data to be coupled from CROME to standard web servers or specialized wireless application servers.

Shown below are actual screen shots of the live wireless "RT variance" application (started by selecting main menu choice "2" in the left screen below).

              

Once the "RT variance" application is selected, a list of all variance reports that have at least one item (typically a sector) is presented to the user.

This list that is initially displayed has the following grammar: <menu choice #> <# items out variance> <network element type> <variance report name>. For example, menu choice "2" shows 39 CELLS (or sectors) out of variance as reported by the CROME report "High Int Blocking".

The above example (middle screen) shows a total of three variance reports that had a positive number of CELLs (i.e. sectors) that exceeded the threshold limits for each of the displayed CROME reports. If no out of variance items were found for a particular report, that report will automatically be excluded from the menu. For this customer, there are five different CELL level variance reports running in this application.

It should be noted that the middle and right screens are the same screen but at different moments in time as the iDEN phone will scroll LINE data back and forth so that a user can read the active line (as marked by the ">" character). This feature allows longer lines to be displayed within the limited screen of the iDEN phone.

We select list choice"2" via the iDEN phones  navigation buttons and the [OK] prompt.  At this point the "RT variance" application will display the first twenty (20) items that exceed the threshold of the "High Int Blocking" report.

              

Each line contains the full name of an EBTS followed by the esmir_cell_id associated with the actual CELL (or sector) that was out of variance.  If you select “INFO” you will get the second and third screens above, showing the complete details of this CELL (type of variance, OMC, BSC, EBTS, SECT (ESMR Cell ID), the Motorola-assigned “network ID” and finally the numeric value of the variance in question -- in this case the I_BCR (Interconnect Block Clear Rate value)

Selecting the [Next N] prompt (where N in this case is 19 or the remaining items in the list) will scroll through the out of variance CELLs (or sectors) until the [Next N] prompt is replaced by the [Done] prompt (the Done prompt indicates no more items).

At this point (or any time previous) the user can "go backward up the HDML stack" to previous screens via the [OK] prompt - or - select the [Done] to return to the main CROME HDML menu.

Countless other modes of operation and key performance statistics can be added to the CROME HDML application suite to meet you specific needs.

To access the HDML CROME pages you will need a HDML/packet data capable iDEN phone and you must direct it to the URL http://wap.quantumsi.com. You can also visit this site via a standard PC Web Browser, and read further information on use of this feature.


Auto Reports and Batch Reporting

 

Auto reports are control templates on the local CROME client host OR on the CROME server that instruct the CROME Client how to automate the process of running one to N CROME Reports.  The ensuing report data can remain on the screen, or optionally have the output of the report go to a files or office tools such as Microsft Excel or even open source office suites such as OpenOffice.

 

Once an Auto Reporting template is configured, they are automatically available for use in the CROME Client.  In addition, the System Administrator can create server Auto Reports that are available to all CROME Users.  Auto reports can also be scheduled by the Unix  “cron” system or the Windows scheduler for fully automatic creation of complex reporting information.

 

Below is a typical auto reporting or batch reporting output (not the multiple work sheets in the Excel report) for a Nortel CDMA system:

 

 

For more information on creating and using Auto Report Files, refer to a separate document, “CROME Auto Reporting Guide”.

Summary

Your organization may not make use of all possible outputs of the CROME system, but just think of the benefits that can be achieved by running one standard generic report and feeding the many different outputs that have been introduced above.

CROME has proven to be highly adaptable and flexible in integrating to other third party tools, allowing the creation of new custom tools, simple publishing to office suites or the Web.

Note if a standard CROME report is updated (or even one formula) all down stream data consumers will reflect the improved reporting situation.  Also due to the object oriented building block nature of the data flow, each new variance report will automatically affect every subsystem: NetExpert, Netcool, the CromeVariance Servlet, and the iDEN phone HDML application.  Each will see the and correctly process the new data or variance report.

Report and Formula Creation Flexibility

CROME allows the user to build up formulas and reports at his or her preferred desktop.  There is not an expensive hard-to-access central server running an operating system the end-user is not skilled with.

The CROME client, which is a JAVA application that runs on every major computing platform and OS, is the only tool needed to design and implement every CROME formula and CROME report.  In addition CROME formula syntax parsers can extract formulas from PDF and WORD documents and automatically create and install "canned" full scale formula sets direct from the system vendors documentation.

Ad-Hoc Formula Creation:

1.        Simple --- a wrapper for any single statistical item measured

2.        Complex --- including mathematical functions and operators

3.        Logical ---  formulas that return 1/0 for true/false based on a threshold function set

4.        Hierarchical --- formulas that consist of combination of other formulas

5.        Constants --- e.g.,  the number of seconds or minutes in the reporting period

Ad-Hoc Report Creation:

1.        Reports are built up with formulas or "raw" primitive data items

2.        Reports can have a default look and feel (i.e. graph or grid)

3.        Reports can have a default sort, limit or threshold

4.        The same reports run at every physical level against every possible group.

Empowerment and Standardization

The system admin only controls the "standard approved and revision controlled reports.”  The end-user is free to clone, modify, or create from scratch any required formula or report at any time to get his or her job done.

With CROME, end users are empowered - there is no developer or corporate data analyst that becomes a bottle neck, yet your organization maintains control of the standard approved report sets via completely traceable reporting items.

Report Generation Flexibility

The key aspect to understanding CROME and its capabilities is that a report is generic and can be applied across any dimension (time, range, filter, or custom group) on any type of network element.

For example, for a typical small Wireless provider having 16 BSCs and one MSC, there are thousands of permutations or different outputs that can be generated from one (1) generic CROME report definition, via either the CROME GUI and/or batch Unix style crontab scripting.

Typically one CROME report will replace four to seven (4-7) full custom reports in any organization where CROME is deployed.  The flexibility of CROME reports is due to advanced system design in which a rich end user filter can be applied to any basic reporting template.   Any existing report and formula can be copied and adjusted for fast alterations or quick "what if" analysis.

The physical dimension:

1.        On a single sector

2.        A custom group of sectors

3.        A single site

4.        A custom group of sites

5.        A single BSC

6.        A custom group of BSCs

7.        A single OMC or OMCR

8.        A custom group of OMCs or OMCRs

The time dimension

1.        Fifteen minute intervals

2.        Half Hour intervals

3.        One hour intervals

4.        Daily intervals

5.        Weekly intervals

6.        Monthly intervals

7.        Yearly intervals

Any definable range (from a half hour to a decade)

Reports are generic and designed to run across any time range.  Consider a Minutes of Use (MOU) report.  The end-user (or automated process) uses the same report to show the half hour data for a single sector as displaying the weekly MOU statistics for all OMCs in a national region.

With or without weekends or holidays

Weekends and holidays can be excluded or forced to NULL/0 to facilitate trending and analysis.  In fact special report filters like "Friday's Only" can be quickly constructed by turning off the other six days of the week and running the report.

Special defined service periods or busy hours, for instantaneous reporting

1.        Multiple service periods / production window (i.e. 6am to 6pm) can be made as per your organizations needs.

2.        Multiple busy hours generated from any formula for the day of analysis, across a sliding business week, or across a fixed set of weekdays.

3.        Multiple busy hours generated from foreign systems i.e. applying the busy hour from one or more DMS switches to OMC RF performance data.

"Busy Hour on the Fly" from any report, for any formula, for any period

1.        For any day, period, or even the across the last 2+ hours of the current day, any report can be run and then used any formula in that report can be selected to "generate" a busy hour view on that formula.

2.        Although not as fast as "Special defined service periods or busy hours, for instantaneous reporting" above, the Busy Hour on the Fly" feature allows non-standard reporting for any specialized  analysis need.

Customized thresholds, sorts and limits

1.        Any value (raw statistical data (pegs) or formula) can be tested against a threshold to limit the amount of data displayed and/or exported.  Any threshold can be made a default.

2.        Any value (peg or formula) can be made the current or default sort (ascending or descending)

3.        Limits can be applied to reports to show only the top "N" (i.e. top 5, top 10, top 20) items.

 

Drill Up, Same, Down

1.        Any report that Views by Element can be “Drilled” to see the same information about sub-elements.  For example, you can run a site (BTS) report and then decide to view the Sector data for any sets of Sites in your report (shown is a drill down for one (1) site):

 

 

                The resulting report shows only those sectors related to the "Drill Down" operation:

 

 

The ability to "Drill" within a data set provides engineers with an extremely powerful method to analyze data sets and quickly determine a "root" cause as well as the scope of particular issues such as configuration management mistakes or RF interference.


2.        Any element or set of elements can be highlighted and then “drilled”.  For example three elements are highlighted below and a "Drill Up" operation is selected to show the BSCs to which selected EBTSs belong.  Note, you can "right mouse click" to bring up a popup menu which functions the same as the "Drill" menu bar.

 

 

The resulting report shows only those sectors related to the "Drill Down" operation (with a "Drill Same" operation applied to show only one i.e. the selected BSC):

 

 

The resulting report (from the action above) would shows only one BSC (no image shown), at which point other "Drill" operations can be applied (i.e. Up, Same, or Down) to quickly navigate the solution space.

 

3.        You can even perform "Drill" operations from the "graph" view of a CROME report item:

 

 

The resulting graph is shown below, with yet another "Drill" operation about to be invoked:

 

Generic, System Independent and Extensible

CROME can be applied to many different types of diverse network elements or system types.  This document highlights ESMIR or iDEN technology, however the CROME software is so generic the same software supports a wide range of telecommunication equipment.

 

In the event that your organization's equipment is not currently supported, depending upon complexity Quantum typically charges between $25K to $50K USD to build custom data interface gateways for a new system type.

iDEN OMC system support

This OMC Reporting capability has been deployed since 1997, featuring reporting capabilities for multiple OMC management devices utilizing direct high speed database replication collection from the OMC into the CROME system.  The current version of CROME supports eight different network elements in the iDEN OMC system: OMC, BSC, EBTS, Sectors, Channels, PCCH, Carriers, and MTL

In addition, selected provisioning data is also maintained in the CROME system to automatically allow extreme analysis functions like Erlang-C analysis in the presence of multiple service types operating under dynamic channel allocation with minimum, maximum, and preferred allocations for each service type.

iDEN DAP system support

This DAP capability has been deployed since 1998, featuring reporting capabilities for multiple Dispatch Application Processors (or DAPs) utilizing direct high speed database replication collection from the DAP into the CROME system.

iDEN MDG system support

This MDG capability has been deployed since 1998, featuring reporting capabilities for MDGs utilizing direct high speed database replication collection from the MDG into the CROME system.

Nortel DMS family system support

This MSC capability has been deployed since 1998, featuring reporting capabilities for MSCs utilizing TCP/IP, X.25 or serial data collection from the Switch into the CROME system.

Nortel GSM OMCR family system support

This OMCR capability has been deployed since 1998, featuring reporting capabilities for OMCRs utilizing FTAM Q3 interfaces for statistical data collection into the CROME system

Nortel DMS-MTX CMDA family system support

This CDMA capability has been deployed since 2006, featuring reporting capabilities for CDMA deployments on a Nortel DMS-MTX CMDA and switch running NBSS 13.0 for statistical data collection into the CROME system


Other TBD System Types

CROME is flexible enough to be extended to other systems WITH NO CHANGES to the Client or the server.  The only requirement is that a specific plug-in gateway be developed for the new system type.

Of course, for every system type Quantum develops the generic mapping rules that allow the CROME system to report with ease on any needed statistical value reported by a given network element (NE) under any given major system type.

Typically performance reporting raw statistical data (i.e. pegs) are generated from and associated with a particular physical layer in a network hierarchy.  For example the low level "peg" set or available statistical measures for an iDEN MDG are shown below:

The above "peg" set will be vastly different than those found on an iDEN OMC or other systems by other system vendors.

Summary

It should be apparent that CROME is very flexible in the wide range of top level systems (and associated sub-elements) it can support.  This is accomplished by a normalized data interface as defined by the CROME open Gateway API that all custom gateways adhere too in order to feed data into its internal RDBMS in a standard generic fashion on the CROME server.


Corporate/National Reporting

Another powerful feature of CROME is its ability to quickly provide management and corporate reports.  CROME’s unique and efficient database organization and flexible user interface make it possible to provide not only low-level engineering analysis, but also high-level corporate summaries.

CROME’s database is organized in a hierarchy of network elements, such that corporate or market data is already “summarized” in the database, making access to this national data extremely fast.  For example, viewing EBTS performance data across multiple OMCs at a bouncing-busy-hour accesses less than 1/100th of the data required by competing systems.

CROME also provides a means to design a custom corporate organization, be it geographical (e.g., by city, county, state) or by management-assigned areas.

Geographical Organizations

For example, one customer assigns “clusters” to appropriate engineers, and the clusters are geographically summarized within cities and states as follows:

 

In the above example the assignments are in fact automatically imported from a foreign database system completely external to Quantum's CROME product.

For this particular market, consisting of the Western United States, clusters have been defined (note Las Vegas is organized as part of the “SoCal” region,  but the customer could just as easily organize, and at any time change, the cluster grouping to be completed separated by state).  They could also maintain multiple trees of different organizations, if so desired.  For example, a local market could organize their reports by engineer-assigned clusters and custom geographical groups, while a corporate manager could organize reports based on city, county, and state.

Any CROME report can optionally run against any corporate or market organization, such as these geographically based clusters.  The following report was run and summarized by each of the folders shown above.  Using standard Microsoft Excel point-and-click functions, we can further summarize based on any grouping.  For example, by changing “level” to “2”, we can see all data summarized by state:

 

The following example takes the same information and, using standard Excel point-and-click functions, shows the detail down to the city or sub-city region:

 

 

Automatic Maintenance of Geographical Groupings

A wireless infrastructure is constantly growing and changing.  For this reason, CROME provides a means to automatically keep geographical data organized for market and national reporting needs.  CROME removes the need for a corporation to constantly manually maintain grouping information every time a site moves or goes on-line.

CROME does this through an Open API for Site Grouping, where a set of simple rules can be defined to trigger automated organization.  For example, since CROME automatically determines the latitude/longitude of each Site, it can automatically build city/county/state organizations.  This means that market and corporate reports can show any and all data summarize by city, county or state.  Another organization may be based upon a pre-defined naming standard, or via a lookup to another database for mappings between site names and engineer-assigned clusters (or some combination of all of those).

Thus, CROME can automate any desired market or corporate organization, such that summarized reporting can be easily and quickly maintained with little effort and little maintenance of the organizational information.

 

CROME Designed for National Scalability

CROME is a distributed client-server based technology, where database management and data summarization is managed on a server or set of servers.  User report processing, which may include some high-computational number crunching, is properly delegated to individual user PCs (though the use of JAVA technology).  In this manner, CROME can easily grow to national or international scales without affecting the speed in which users can generate reports.

This is an important distinction between CROME and other PM solutions -- more data does not mean slower reporting.  CROME actually has the ability to run more efficiently in a nationally-scaled system, because its design can make use of multiple servers as replicated database repositories for instantaneous hot backup.  For example, CROME can be organized to store the same data (or most recent data) on more than one server, which requires more disk space but provides a means to circumvent downtime in the event of equipment or network failures, or scheduled maintenance.  CROME was designed to explicitly manage “hot failover” databases, where the user is completely unaware that the software has transparently shifted to a database on another backup server -- it simply works.

 


Advanced Analysis Features

CROME can be used to perform many advanced analysis of the elements in your network including but not limited to:

Migration Tracking

MIGRATON, finding and properly tracking network elements that migrate across time or undergo name changes:

CROME can track sites (EBTSs) and their associated sectors as they are migrated across base station controllers (BSCs).  CROME is the only PM today system that is capable of this level of low level migration tracking and filtering.

When reporting on a specific site (or EBTS) CROME will accurately follow the site even if it is moved across several BSCs over the range of the Query.

When reporting on a specific BSC, CROME will accurately include a migrating EBTS for only those days that the BSC owned the migrating site.

Grouping

GROUPING, is the process of a system administrator or end-user creating an add-hoc set or group via individual items (or network elements) to make custom sets (or clusters) to report against.

The CROME system allows both traditional (i.e. element based) grouping as described above.  In addition CROME supports advanced pattern match based grouping which simplifies the maintenance of cluster sets through the application of regular expressions (REs) against existing names.

Shown below is a traditional group (i.e. individually selected items):

One of the downsides of traditional groups is that if an element (such as 2TS5SIG/2W) is renamed (or migrated) the group will become "stale" and out of date.  Any report run after this change will typically fail to find the item.

Quantum's exclusive regular expression pattern matching solves the above problem and can be used to create "always current" groups. Shown below is a special "US state based RE set" that will always extract all Arizona EBTS sites from the Phoenix OMC.

No matter how many new sites are added the above RE will always include any EBTS that begins with the two characters "AZ" thus if the naming standard is always followed the group is always current.

Note if Arizona sites are served by three OMC's say phx01 (as above) phx02 (future?) and tex02 (future?) the RE-set can be updated as follows:

Shown below is a special "route based RE set" where trunks are named based upon the function and destination of the route.  This set of RE set is currently used in a CROME deployment in Taiwan that comprises eight (8) Nortel MSC switches.

The symbols used by CROME in the group editor are as follows:

            white      Item available for selection or item included in a group

            red          Item excluded from a group

            pink        Item in group but is stale, i.e. this item doesn't exist on the current date.

            white      RE in group as denoted by the letter 'R'

            red          RE is excluded from group as denoted by the letter 'R'

Thus in the above "route based RE set" example we first include everything than we exclude specific patterns, in this fashion very complex RE-sets can be constructed to meet nearly any organizational need.

 

GIS reporting

In addition to the CROME real-time variance reporting (i.e. a distributed multi-user big screen display) described earlier, CROME can also display all ad-hoc report results in a map based GIS view when the elements reported on can map into a geographic location.

Selecting the "Geo Map" menu item when available will invoke a highly versatile GIS data explorer to analyze the data from any CROME report.

The Geo Reporting feature of CROME allow the analyst to interactively alter thresh holds for all formulas in a given report to determine visually faults or congestion in the network infrastructure.

This feature is very similar to the CROME real-time variance reporting, but allows the end user to interactively adjust thresh holds and included items.

 

The Geo Reporting feature allows for a high degree of interaction and visual display options such as suppressing site display by value thresh holds, and also gradient coloring by value or value range as shown above.

 

Neighbor Cells and Airgen Worksheets

The CROME Site Mapping Geographic tool integrates Site Configuration information with two powerful features: Show Neighbor Cells and Airgen Worksheets

 

Neighbor Cells

Clicking the right mouse button over any site on the map and choosing "Show Neighbor Cells" launches a powerful feature to analyze the "neighbors" of sectors within the site.

A "Neighbor Cell Window" will appear, showing the site name and, for each sector in the site, a list of each site/sector that is configured as a "neighbor" for the sector:

Note in the picture above that purple lines appear on the map. These lines link the site (in the center) with all of its neighbor sites.

The Neighbor Cell Window contains, underneath the words SECTOR 1, SECTOR 2, and SECTOR 3, "check-boxes" labeled "Show links on map". If you un-check any of these boxes, the purple lines on the screen which represent links for that sector will disappear on the map (and of course re-appear when you check the box back "on").

There are three ways to switch the Neighbor Cell Window to view a different site. One way is to double-click on any of the sites listed in the window. This will cause the Neighbor Cell Window to load and display the neighbors for the selected site, and consequently the "purple lines" will shift to the new site.  The second way is simply to click the right-mouse button on a new site and choose "Show Neighbor Cells".

 

The Neighbor Cell Window also contains, as shown below, a pull-down choice of all previous sites that have been displayed in the Neighbor Cell Window since the map was first brought up. This is the third way to switch the window to a different site. You can always "jump back" to a previously displayed site by finding it in this list:

Airgen Worksheets

The "Show Worksheet" menu choice launches a new web browse page which contains the current Airgen Worksheet for this site. The web page contains all parameters, plus a number of synthetic parameters created in CROME (these parameters all begin with "0_" and contains useful information like BSC name, Lat/Lon, etc.).

A sample web page is shown below. This picture only shows the top of the scrolled page -- there is much more below on the actual page. Note that "EBTS" is highlighted, as are all sub "elements" further down the page (e.g., CELL, CARR, BR, NEIG, etc.) so that it is clear which parameters belong to which level.

Since the Airgen worksheets are quite long, custom Airgen "reports" can be very quickly created to show useful subsets of the data, making it easier to quickly get the desired information.

Three of these custom reports have already been created, which are listed under the menu item "Worksheet Reports":

The "RF Neighbor Report" isolates pertinent information about the CELLS and their neighbors. "RF BR Info Report" focuses on base radios, and "EBTS Identification Report" gives all of the CROME-created synthetic parameters, as well as some airgen parameters, that help identify this site.

Below is an example of the "RF Neighbor Report" that appears in the web page (note again that most of the report is scrolled off the screen):

 


GIS reporting by arbitrary area (GIS Grouping)

Another valuable feature of CROME's GIS reporting is the ability to create user definable polygons via a CROME GIS group editor to define arbitrary boundaries and have CROME automatically group elements that can map into a geographic location for reporting purposes.

               

 

A few geo-coded groupings that have been created in CROME by customers include:

·         zip or postal codes,

·         state or county boundaries,

·         special service areas

·         responsible engineer viewpoints

Like any Group created in the CROME system it can be install system wide for all users or even created locally and used locally by any end user running the CROME GUI. 

In addition any existing (or new) CROME report that is run in a non-GIS fashion can also be applied to a geo-coded grouping  as long as selects elements that have a geo location.

The CROME Group Editor is shown below with a simple geo-coded group, which was used to “lasso” a set of sites and run a report against the selected area.

The cross integration of CROME’s built in GIS and data mining capabilities is incredibly easy to use and extremely powerful and can not be matched by other PM tool vendors.

Not only can CROME easily geo-code polygon areas but these areas can be used with the non-GIS exact element selection and enhanced regular expression “Grouping” modes of CROME.

Extendable to non-PM domains (e.g. Drive Test Analysis)

Quantum has even used the CROME suite to support CDMA and iDEN drive test data analysis this is another product line, but it highlights the flexibility of the CROME solution.  Below we show CROME configured to display and analyze iDEN data (with RSSI, SQE, and FREQ):

In the case of drive test analysis typically we do not want to aggregate data by time.

 

Selecting CROME’s GeoMap tool we can see the test data properly plotted

 

 

Of course you can switch between any formulas in your ad-hoc report

 

CROME provides special spatial aggregation functions to analyze both drive test and PM data into normalized regions.  This  “GRID CELL” function allows yet another style of data reduction. 

 

Below the “maximum” RSSI value of drive test data being spatially aggregated into 500 meter regions.

 

 

The same “GRID CELL” technique can be applied to PM based sector data, for example the below report is used to first determine high use regions and then analyze the measurements in each region via a mouse over.

 

 

Once again you can switch between any formulas in your ad-hoc report

 


Distribution Graphs (e.g. the percent of items):

 

CROME provides a graph display which shows a distribution (e.g. the percent of items) that are at or below a given value.   The below graph is always available via the menu option:

 

 

Full interpretive Scripting via Plug-In Support


Power User's can write JAVA based scripts to do any function they desire with CROME data and interact with the external world in any fashion they see fit -  an example correlation script which creates a new  CROME report from an existing report is shown below:

 

// standard initialization stuff:
CPluginData data = (CPluginData)S_pluginHash.get("cromedata");
CPluginDataRow [] rows = data.getRows();
CPluginResult result = new CPluginResult();

// initialize a Vector of new rows we're going to create
Vector vRows = new Vector();

// find the column numbers for Event Code and EcIo
int eventCodeColumn = data.getColumnIndex("Event Code");
int ecioColumn = data.getColumnIndex("EcIo (dB)");

// for each row of data
for (int i = 0; i < data.getNumRows(); i++) {

    // if we have a "211" Event code in this row
    if (rows[i].isValidAndEqualTo(eventCodeColumn, 211)) {

        // if this row is already matched up with data (i.e., has a valid EcIo)
        if (rows[i].isValid(ecioColumn)) {

            // if the EcIo value is less than 5.0
            if (rows[j].getValid(ecioColumn) < 5.0) {

                // this row is good as is -- include it
                vRows.addElement(rows[i]);
            }
        }

       *
       *
       *

 

Of course since it is JAVA the end user could even implement a web server inside a CROME report, send a CROME generated graph directly to his mobile phone, or pretty print a report to a PDF file and E-mail it to a distribution list (how's that for flexibility).  Since the scripts operate "client side" they can not impact or load the main CROME server(s) with extensive unforeseen processing requirements.

Erlang and GOS support

For "On the Fly" ERLANG and Grade of Service (GOS) support, the CROME application calculates and applies Erlang functions and Grade of Service functions on demand to eliminate the use of manual table lookups.

Please refer to the section on "Resource Analysis and Capacity Planning" in this document where two fairly detailed examples are shown.

Note CROME is extensible in that it can support other distributions and equations, for the purposes of clarity the some of the built-in Erlang-B and Erlang-C base mathematics are described below.

Erlang B

The Erlang B formula is based on the assumptions that there are infinite sources, blocked calls are cleared, and constant or exponentially distributed holding times. 

where

A=Total traffic offered in Erlang

N=Number of servers in a full availability group

P=Probability of loss

Erlang C

The Erlang C model is based on following assumptions: infinite sources, blocked calls are queued, and holding times are exponentially distributed.  Another assumption is that waiting calls are served in the order in which they arrive in the waiting line (first-in-first-out).  Exponentially distributed means that the number of calls having increasingly longer duration is proportionately decreasing.  This means that the number of people making longer calls (e.g., 30 minutes) is much smaller than the number of people making shorter calls (e.g., 5 minutes).

In queuing systems, several terms are used to designate  the grade of service.

Average delay of all calls -- Indicates the average time that offered calls must wait for a connection to be established.  This includes those calls that have zero delays.

Average delay of delayed calls -- Indicates the average time that delayed calls (delay greater than zero) must wait for a connection to be established.

Probability of delay -- Indicates the probability that an offered call is not handled immediately.

Probability of delay being exceeded -- Indicates the probability that an offered call must wait longer than a specified time for establishment of a concern.

where

A=Total traffic offered in Erlangs

N=Number of servers in a full availability group

P(>0)=Probability of delay greater than 0

The probability of delay greater than t is:

The average delay, D1, on all calls is:

The average delay, D2, on calls delayed is

Heterogeneous System Support

Flexible system support, CROME does not require that all top level systems be homogeneous (i.e. at the same system revision) and will not only report on the differences (i.e. missing pegs or uniformity of schemas) but will also generate correct reports in such situations.

In large deployments it is unlikely and sometimes rare that all available statistical sets are identical. Unlike other PM systems that simply stop working until all selected sources are identical in the data schema, CROME allows statistical reporting to continue in mixed data sets environments taking advantage of every piece of available data and always reporting the correct results.

For example the below information available from the "View/Schema Info" menu shows that two of the four database tables used to generate the report are missing in one database, swo_05.

The important point is that CROME will continue to run and generate correct results even in the event that you are reporting across mismatched system software from your equipment vendor.  The formulas that can not complete due to a missing table or column simply resolve to blank or null data in the CROME report.  Of course if all the reports formulas use pegs from missing tables your report will be completely empty.

Scalable

CROME was implemented from the ground up as a scalable distributed system.  Unlike other centralized solutions as your business grows to national or even international scope CROME will easily scale (at much better than the typical best case linear performance trend line!) and maintain quick responsive reporting speeds across tens of thousands of network elements (NEs).  This is accomplished via the following design tenants:

1.        Each CROME serves runs independent and in parallel to support scaling via parallel operation.

2.        Only the minimum amount of data that can satisfy any given report is queried from a CROME server.

3.        Each CROME clients perform all intensive mathematical calculations locally to minimize server loading.

4.        The server only server data it does not perform expensive summaries or number crunching.

As a timing example, a typical full custom CROME report showing various minutes of use (MOU) metrics at the sector level run against 1/8th of all subscribers (5,181 sectors) for this iDEN customer takes a mere 9.4 seconds.  Due to adaptive caching algorithms in the CROME client application the same (or similar) reports will run even faster the second or third time (i.e. 6 seconds).  Contrast this typical performance to the number to the hours required by our competitor’s product to produce a similar ad-hoc report.

As your network grows another CROME server can be added to handle a portion (typically 1/N where N is the number of CROME servers) of the overall load.  When a report requires more than one CROME server all data from only those CROME servers required to satisfy a report are queried and fetched in an optimal space/time efficient parallel fashion by the CROME client application.  The bottom line is that CROME can handle massive network infrastructures and unlimited future growth.

Fault Tolerant for 100% uptime

CROME system deployments can be configured to be fully redundant and fault tolerant, in the event of failure another CROME server will seamlessly and transparently take over and all client requests will be satisfied with at most a fifteen (15) second delay to work around the failure.

Of course fault tolerance has an associated cost, to create a high availability CROME additional hardware and storage infrastructure is required to complete a circular mirror or "CROME ring" of all your company’s vital performance reporting data.

Zero Client Administration

Installation of the CROME application

For optimum performance and functionality, the full CROME application can be installed with a single click from a Web page (an image from such a page is shown below).  Note when running as an application, CROME requires a Java Virtual Machine (JVM) to be locally installed on your desktop computer.

Thus for a Win95/98/2000/NT/XP installation you will need to first install a JVM (which provides a runtime base for many different programs much like visual basic) and next install the actual CROME application, both steps are accomplished via wizards started by a simple web click.  The installer automatically finds the "best" JVM available to ensure maximum performance.

Upon installation, CROME is fully integrated into your work environment. For example, a standard startup icon is added to your “Programs” menu on a Win95/98/2000/NT/XP platform.

In addition to the "Crome" launch icon a set of utilities, like uninstall and debug, for the CROME client application under the folder "Programs/QuantumSI/Crome".

Because the output of CROME reports are sent to the default registered viewers in a Win95/98/2000/NT/XP environment, once CROME is installed you can start mining data to your specifications and seamlessly transfer the results into office productivity tools such as Microsoft Excel or PowerPoint via bridging capability built into the CROME client.

NetworkUpdate

CROME contains an extremely useful feature for maintaining the latest version of your software: it automatically updates itself over the network whenever it is out of date! 

If CROME determines that the software you are running (i.e. CROME itself or any related support file) is out of date, your CROME client will notify the user with the following pop-up screen. At this time you will have the option to “Update Now” or  “Skip Update”.

This ensures that you are always running the latest version of the CROME client software (however you have the option to defer updates if you are in a location with a slow internet link).

Built in Help and Support Email Gateway

The CROME system has a complete built in help suite including: pop-up help, a online HTML documentation set, a admin "message of the day" a Email support facility that allows all information of your CROME installation and your current report (if you have an issue with a report) to be sent directly to Quantum engineers.

The highly integrated "Support Email Gateway" enables Quantum to provide extremely quick advise to specific questions and/or problem replication and resolution for complex reports.


Total System Flexibility

 

Most tool suites are designed for one particular task, in contrast to the traditional software design philosophy, QuantumSI values generic extensible code from which our customers gain enormous benefit.  The flexibility and capabilities in CROME are unparalleled.

With this in mind QuantumSI challenges potential customers to ask themselves (and other venders) some serious questions about regarding the flexibility and extensibility of alternate Performance Management (PM) solutions :

·         Can your PM tool analyze different technologies such as iDEN, GSM, and CDMA wireless and wireline data from different switches and vendors in the same application - is it flexible enough to quickly add support for new data sets (i.e. new elements for new services ) ?

·         Can your PM tool process thousands of sectors (or trunks, etc.) per second in a full ad-hoc reporting mode?

·         Can your PM tool support anyone in the enterprise using only a common web browser like IE and Netscape on platforms ranging from PCs to Unix/Linux workstations?

·         Can your PM tool run with different software revisions and mismatched data sets coming from the myriad of network elements in a national infrastructure?

·         Can your PM tool scale from a market to a nation without losing performance and becoming sluggish?

·         Was your PM solution designed from the ground up to be transparently fault tolerant?

·         Is your PM tool flexible enough to analyze other data for example billing data i.e. mining CDR records and even drive test data i.e. tacking a single mobile phone across time and displaying statistical information in an interactive GIS display as it moved?

·         Can your PM tool match CROME's proven near-term and long-term capabilities, CROME not only processes data in real-time but also stores enough data online to perform historical trend analysis via ad-hoc reporting across 5, 10 and even 15 years.

QuantumSI's CROME solution easily meets the challenges set forth above. It is no small wonder that our competitors have been attempting to copy the QuantumSI CROME product suite for years (using our product as the "gold" standard) in an attempt to emulate CROME's technology, feature sets, high level of integration, and ease of use.


Testimonials for CROME by CROME Users

 

CROME has shown itself to be the most dependable tool available today for traffic and Uplink statistical data. ... more rapid and user friendly than cellsight .... To stop using CROME would reduce our ability to quickly identify and attend to service affecting issues. I would be more than willing to testify of the need to keep CROME in this market.

- Rod Woodfield, Nextel Senior RF Engineer

 

I find [Crome] EXTREMELY useful on a daily basis for tracking and isolating dropped calls.  ...  Crome helps identify those [dropped call] times and allows us to be there when it happens.  ... useful for monitoring very recent, datafill parameter and frequency changes or base radio additions without waiting for cellsight data the next day.  ... I expend MUCH less time waiting for crome reports than cellsight data.

- Peter J. Phalen, Nextel Cluster 14 RF Engineer

 

... CROME is a great tool and equips every Nextel Engineer with the ability to better serve our customer base efficiently and effectively. ... [CROME] data was always available.  CellSight would make you wait 24 hours before you can see if a change helped decrease your drop call rate - and the data was unavailable.

- Eric Vanderplas, Nextel RF Engineer

 

I am responsible for all traffic reports to be availed by cluster engineers, managers, and directors.  Having the CROME reporting application installed on my desktop has enable much more flexibility, dependability, and efficiency in running present day reports than CellSIGHT. [In addition] I reference [CROME Real-time Variances] at least once an hour. There have been countless problems that I have seen early enough in the day that we were able to fix before the issue became critical.

- Gabriel Marinelli, Nextel RF Traffic Engineer

 

I had an issue just last week where a trouble ticket came in for a site which had no interconnect.  Looking at [CROME’s] Geo Map, I quickly spotted five more sites which had a similar issue due to a DS3 migration problem the night before. Without this "bird's eye" view, we might not have uncovered the extent and nature of the outage, nor have resolved it so quickly.  I previously used CellSight and was extremely frustrated with the lack of speed and outright failure to pull useful data.  Crome greatly helps us identify hourly trends and variances from day-to-day when troubleshooting dropped call problems.

- Jeff Allen, Nextel RF Engineer

 

The wealth of information that CROME provides is priceless.  I hope Nextel continues to keep CROME operational.

- Michael Downs, Nextel RF Engineer

 

The entire Network Operations Team relies heavily on Crome to compile our Southwest Network Operations Review. Crome is responsible for 99% of the reports generated for this review presented to high level management. ...

 

CROME blows away any competition. .... CROME has given us a information system we can rely on. ...

 

[QSI Support:] Whenever I have brought a question to the attention of Quantum, I have received immediate response. I have even called in the middle of lunch before, and received quantum's attention for over an hour.

- Munjeet Singh, Nextel Network Operations Team

 

I think that crome/wap service [viewing Crome data directly on the iDen Phone] has added a valuable tool in helping me solve problems, particularly the "low ini" tool, it has helped us find problem antennas etc.

 

- Joe Barron, Nextel RF Engineer


Testimonials for CROME by CROME Users (cont.)

 

CROME has proven to be an invaluable tool that is surprisingly powerful and highly intuitive. Within minutes of being exposed to CROME, engineers have been able to view traffic patterns and root out problems in a more sophisticated and expedient manner then ever before.

 

CROME has given us the ability to engage less experienced and non-technical people into the realm of performance and fault management. Field tech[nition]s use the wireless applications interface to detect problems while in the field. Network monitoring centers now have the ability to react to problems even when no alarms are detected. Customer care operators have a birds-eye-view of site status on their desk-top PC's.

 

During the 2002 winter Olympics, CROME allowed our engineers and managers to keep a real-time view of network capacity and performance. All aspects of the network stayed under great scrutiny before, during, and immediately after the Olympics. This allowed our team to react to any unexpected traffic patterns or outages immediately had they occurred.

 

In my opinion, CROME gets the gold for being the most accurate, responsive, and reliable performance reporting tool in the industry.

- Ron Schneirson, SR. Manager of Network Engineering (West).

 


Testimonials for QSI Support by QSI Customers

 

To Whom It May Concern,

 

The QSI Crome tool has been an absolute necessity. As good as a tool can be, it means nothing if you don't have the technical support to assist you with different tasks. The QSI team has been phenomenal when it comes to supporting the Network Operations team in Southern California and Southern Nevada. I personally have called on QSI at awkward times in the past and received developer support. One night, at 22:00 I called their offices because I wanted to use Crome to help determine the detrimental effects of an OMC outage. I expected to receive voicemail, but instead I was on the line with one of the developers who had forwarded the office phone to his home phone "just in case" we needed assistance.

 

Countless times QSI has given me a heads up on statistical issues with the OMC. Issues that would have gone un-noticed if it was not for them bringing it to my attention. Sometimes one of the Employees in Engineering catches the statistical outage, but %100 of the time they notice it by using Crome. If I have to go back and manually re-parse the statistics, I can always count on QSI to reload the data from the OMC so that Crome does not contain a data outage (which has a huge effect on our Traffic Engineering team).

 

Everyone I have spoken with in Southern California agrees with my statements above. Working with QSI has been an awesome experience. If we could get Motorola to support us in this fashion, we just might be able to reach the elusive "Five Nines"....

 

Munjeet P. Singh

MSO Engineer

(949)-795-1331

Munjeet.Singh@Nextel.com 

 

 


Testimonials for QSI Support by QSI Customers (cont.)

 

To Whom It May Concern:

 

I would like to relate to you my personal experience working with QSI as a vendor. My relationship with QSI goes back to 1996. At that time I took over the administration of the DOME, NetExpert, and MCB servers in the Atlanta MSO. Over the years I have had many occasions to call and work with QSI. In nearly every case I was able to get someone to help me the first time I called. On the rare occasions when I did not get someone on the phone I always had a response within a few hours. The technical knowledge of the QSI staff on their products is the best I have seen from any vendor I have dealt with.

 

The worst case scenario with a DOME server is to have a new SR release that changes the performance database structure. I have personally been through several of these and QSI has never failed to keep the DOME server up and running. In these cases the only degradation in service was in relation to the new tables or fields introduced into the OMC database. This is a testament to the robust design of the system. The service rendered by QSI in these cases was always prompt, professional, and well executed. From the time I notified QSI I never had to wait more than about 48 hours to have all of the new information in the system. This includes making all of the changes to the database structure, the system programming, and loading the new data back to the date of the SR cut. This was done in cases where QSI had no prior knowledge of the date of the SR upgrade or any documentation on the changes made.

 

Overall the service I have received from QSI has been quite good. If you have any questions feel free to contact me.

 

Sincerely

David D. McKinny

Regional MSO Engineering Manager

Southeast Region

770-326-7964 (desk)

404-557-9208 (iden)

Dave.McKinny@Nextel.com


Testimonials for QSI Support by QSI Customers (cont.)

 

To Whom It May Concern:

 

QSI is a solutions provider which NEXTEL have a long and great relationship with.  They have been a vendor which work with the customers to best satisfy the customers.  Quantum Systems Integrators have been providing exceptional support in following manner:

 

1 Professional Customer Support - Unlike many of the vendors that NEXTEL deals with, QSI have been providing NEXTEL with professional support focusing on 100% customer satisfaction.  QSI have been very responsive and timely when it comes to supporting their customers.  They listen to customer requirement and meet the expectation of the customer's needs.

 

2 Understanding NEXTEL's unique Business Need - With the extensive working relationship with NEXTEL, they have advantage of providing best solutions for NEXTEL.  Based on some of the successful project results in the past, QSI proved provided solutions that best addressed NEXTEL's business needs. Also, their expertise in IDEN brings innovated solutions tailored just right for NEXTEL.

 

3 Quality Software -  QSI's software is already deployed in NEXTEL's nation-wide network providing fundamental infrastructure to manage complex IDEN network because of their superb quality of the software. Some of these applications include MCB, Netra Servers, DOME, CROME, etc. It is fair to say that our company wide NMS solutions is completely depends on QSI's solutions.

 

 

Young Cho

NEXTEL Communications

Lab Operations Manager - NEXTEL Technology Lab

Office  (703) 762-7789

iDen  (703) 906-3457

e-mail   young.cho@nextel.com

 


Testimonials for QSI Support by QSI Customers (cont.)

 

To whom it may concern:

 

I would like to express my support for QSI and its staff.  I have dealt with them on a weekly basis since 1998.  Their product has been a tremendous help to Nextel.  QSI has always been quick to give support.  The staff is highly knowledgeable and easy to work with.  I know that my knowledge of Unix and Sun servers has greatly increased through working with the QSI staff.  I wish they had taught my Unix class at college. 

 

QSI has a genuine desire to see their customers needs met.  Nextel can only  benefit from a relationship with such a technically strong and professional company.

 

Sincerely,

Amy F. Hemphill

Network Engineer II

Nextel Communications

770-326-7948

Amy.Hemphill@nextel.com


Customer supplied "Cost Benefit" results

 

TASK DESCRIPTION

Task Time in Minutes by department or major activity

Audit Results

 

Savings

 

 

 

 

 

 

 

 

 

 

 

CROME vs. IN HOUSE MULTI-MILLION DOLLAR LEGACY SYSTEM

RF

NOC

MSO ENG

PSTN TRAFFIC

NET OPT.

CAP. ENG

TSG

Total Minutes spent using CROME

Total Minutes spent using Legacy System

Total Minutes Saved

Accessible/Ease of Use (PC or Unix, Fast)

2.5 

2.5 

2.5 

2.5 

2.5 

2.5 

2.5 

17.5 

420.0 

402.5 

Slight/Major Network Mod Impact- Hourly same day hourly stats

1.0 

1.0 

1.0 

1.0 

1.0 

 

1.0 

6.0 

360.0 

354.0 

Customer Problem Tied to Network Issues

5.0 

5.0 

5.0 

5.0 

 

 

5.0 

25.0 

900.0 

875.0 

Short and Longterm Trends

 

 

2.0 

2.0 

2.0 

2.0 

 

8.0 

240.0 

232.0 

Customer on singular sector - Focused reporting

3.0 

3.0 

3.0 

 

 

 

3.0 

12.0 

240.0 

228.0 

Hardware Problems

10.0 

10.0 

10.0 

 

 

 

10.0 

40.0 

240.0 

200.0 

Site Degradation Analysis

10.0 

10.0 

10.0 

 

 

 

10.0 

40.0 

240.0 

200.0 

Same Day Hourly Stats (FNE/RF)

1.0 

1.0 

1.0 

 

 

 

 

3.0 

45.0 

42.0 

ADD-HOC Queries for trouble shooting

10.0 

 

10.0 

10.0 

 

 

 

30.0 

360.0 

330.0 

Hour By Hour per Sector Stats

1.0 

1.0 

 

 

 

 

1.0 

3.0 

45.0 

42.0 

Freq. Interference Problems

1.0 

1.0 

 

 

 

 

1.0 

3.0 

90.0 

87.0 

Proactive Repairs

10.0 

10.0 

10.0 

 

 

 

 

30.0 

180.0 

150.0 

Wireless Application (WAP) Interface In the Field

1.0 

1.0 

 

 

 

 

 

2.0 

100.0 

98.0 

NE Degradation Analysis

 

 

10.0 

 

10.0 

 

 

20.0 

120.0 

100.0 

No EBTS Traffic (Disp or Inter) or other user defined problems Mapped, and NetExpert, or NetCool (PM to FM)

0.5 

0.5 

 

 

 

 

0.5 

1.5 

720.0 

718.5 

Recent Find (4/17/01) - Sect split, no traffic. Tied to EBTS/BSC Load Mismatch

20.0 

 

 

 

 

 

 

20.0 

120.0 

100.0 

 

 

 

 

WEEKLY LOCAL ENG MINUTES

261.0 min.

4,420.0 min.

4,159.0 min.

 

 

 

 

WEEKLY LOCAL ENG HOURS

4.4 hr. 

73.7 hr. 

69.3 hr. 

                                                                                                                                                                                          

Actual:     70.0 hours saved per week (4 OMC’s in Southern California), approx. 2 Associates     = savings of      $240,000 USD/year

Est.:       157.5 hours saved per week (9 OMC’s Western Region), approx. 4 Associates              = savings of      $480,000 USD/year

Est.:       787.5 hours saved per week (63 OMC’s Full Nation USA), approx. 20 Associates          = savings of   $2,400,000 USD/year


 These savings only look at actual time spent in report preparation and do not reflect other “cost benefits” such as improved accuracy, improved uptime, increased ability to perform proactive corrections via real time PM analysis and full ad-hoc reporting. The full burden rate for a Nextel USA associate is approximately $120,000/year inclusive of salary, benefits and other overhead costs. 


CROME Performance and Scaling (iDEN OMC) - Benchmark run on June 12, 2002

 

Rpt

Report

number of DBs, or iDEN OMCs

number of DB tables

number of DB table columns

number of report formulas

number of elements (sectors)

DB connect time in sec.

client compute time in sec.

total report time in sec.

report time / #SECT in sec.

report time / #OMCs in sec.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Minutes of Use (int/disp/tot) at an element bouncing busy hour

 

 

 

 

 

 

 

 

 

 

#1

MOU/SECT/pkt BBH

6

1

5

4

5181

5.40

+

1.00

=

6.40

 

0.0012

1.0667

#1

MOU/SECT/pkt BBH

5

1

5

4

4532

4.20

+

0.80

=

5.00

 

0.0011

1.0000

#1

MOU/SECT/pkt BBH

4

1

5

4

3944

3.70

+

0.70

=

4.40

 

0.0011

1.1000

#1

MOU/SECT/pkt BBH

3

1

5

4

3067

2.90

+

0.60

=

3.50

 

0.0011

1.1667

#1

MOU/SECT/pkt BBH

2

1

5

4

1982

1.80

+

0.30

=

2.10

 

0.0011

1.0500

#1

MOU/SECT/pkt BBH

1

1

5

4

1036

1.10

+

0.20

=

1.30

 

0.0013

1.3000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Dispatch Erlang Analysis at an element bouncing busy hour

 

 

 

 

 

 

 

 

 

 

#2

ERL/SECT/disp BBH

6

3

31

24

5,181

17.2

+

6.7

=

23.9

 

0.0046

3.9833

#2

ERL/SECT/disp BBH

5

3

31

24

4,532

14.1

+

5.8

=

19.9

 

0.0044

3.9800

#2

ERL/SECT/disp BBH

4

3

31

24

3,944

12.4

+

5.2

=

17.6

 

0.0045

4.4000

#2

ERL/SECT/disp BBH

3

3

31

24

3,067

9.1

+

4.1

=

13.2

 

0.0043

4.4000

#2

ERL/SECT/disp BBH

2

3

31

24

1,982

6.2

+

2.2

=

8.4

 

0.0042

4.2000

#2

ERL/SECT/disp BBH

1

3

31

24

1036

3.2

+

1.1

=

4.3

 

0.0042

4.3000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Once any report is run in a given month CROME caches the cross reference data for improved reporting speeds of all

 

Databases holds

subsequent reports run at that network level (or above), regardless of the formulas used.  For this reason Run's 2-N are

 

900+ days of data

somewhat faster than the first report.  The above table results use cached cross reference data (i.e. runs 2-N).

 

 

 

report runs are

 

 

 

 

 

 

 

 

 

 

 

 

 

for one day only

The Client PC used was a 1.33GHz Athlon ($450 USD) with a Sun Microsystems 1.3.0_002 Java VM.  The Client PC and the

 

 

CROME server were physically separated, communicating via  fractional-T1 with 19 DSOs (I.e. 1214 mbps)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The two benchmarks reports are typical examples of a simple ad-hoc report #1 (MON/SECT/pkt BBH) that accesses a few

 

 

 

columns in one table and a complex ad-hoc report #2 (ERL/SECT/disp BBH) that access many columns in multiple tables

 

 

 

and performs extensive mathematical analysis (on the client PC).

 

 

 

 

 

 

 

 

 

 

 

 


CROME Performance and Scaling (Nortel MSC) - Benchmark run on June 12, 2002

 

Rpt

Report

number of DBs, or Nortel MSCs

number of DB tables

number of DB table columns

number of report formulas

number of elements (TRKs)

DB connect time in sec.

client compute time in sec.

total report time in sec.

report time / #TRK in sec.

report time / #MSCs in sec.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Erlang B/Grade of Service for number of justified DS0s on each TRK group

 

 

 

 

 

 

 

 

 

 

#1

RTR_NEED_DS0/TRK/pkt BBH

12

1

3

6

1141

0.90

+

0.40

=

1.30

 

0.0011

0.1083

#1

RTR_NEED_DS0/TRK/pkt BBH

11

1

3

6

1008

0.60

+

0.40

=

1.00

 

0.0010

0.0909

#1

RTR_NEED_DS0/TRK/pkt BBH

10

1

3

6

891

0.80

+

0.30

=

1.10

 

0.0012

0.1100

#1

RTR_NEED_DS0/TRK/pkt BBH

7

1

3

6

669

0.50

+

0.20

=

0.70

 

0.0010

0.1000

#1

RTR_NEED_DS0/TRK/pkt BBH

6

1

3

6

583

0.50

+

0.20

=

0.70

 

0.0012

0.1167

#1

RTR_NEED_DS0/TRK/pkt BBH

5

1

3

6

508

0.40

+

0.20

=

0.60

 

0.0012

0.1200

#1

RTR_NEED_DS0/TRK/pkt BBH

4

1

3

6

410

0.60

+

0.20

=

0.80

 

0.0020

0.2000

#1

RTR_NEED_DS0/TRK/pkt BBH

3

1

3

6

318

0.50

+

0.10

=

0.60

 

0.0019

0.2000

#1

RTR_NEED_DS0/TRK/pkt BBH

2

1

3

6

213

0.20

+

0.10

=

0.30

 

0.0014

0.1500

#1

RTR_NEED_DS0/TRK/pkt BBH

1

1

3

6

103

0.20

+

0.10

=

0.30

 

0.0029

0.3000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Once any report is run in a given month CROME caches the cross reference data for improved reporting speeds of all subsequent reports run at that network level (or above), regardless of the formulas used.  For this reason Run's 2-N are somewhat faster than the first report.  The above table results use cached cross reference data (i.e. runs 2-N).  The Client PC used was a 1.33GHz Athlon ($450 USD) with a Sun Microsystems 1.3.0_002 Java VM.  The Client PC and the CROME server were physically separated, communicating via  fractional-T1 with 19 DS0s (I.e. 1214 mbps)

This benchmark reports is a typical examples of a medium complexity ad-hoc report #1 (RTR_NEED_DS0/TRK/pkt BBH) that access many columns a single table and performs extensive mathematical analysis (on the client PC).  The actual report will determine the number of DS0's that should be added to any out of variance TRUNK group to meet a 1% GOS using CROME's Erlang-B built in analysis capabilities.

The following results from (RTR_NEED_DS0/TRK/pkt BBH) across ten (10) MSCs were obtained in 1.1 seconds the report displays the four (4) TRK groups (out of 891 total) that fail to meet a grade of service (GOS) threshold of 1% and calculates the required number of resources (i.e. justified DS0s) needed to meet the target threshold (I.e. 1%) for each particular route.

MSC

TRK

trk#

BBH

WCCT

JUSTIFIED

CARRIED

CARRIED

CARRIED TRU

 

 

 

using TRU

 

DS0s

TRU GOS

TRU CCS

ERLANGS

LNBHCAXKCM0

ANHMCA0295T/2W

90

15:30

768

793

2.17%

27337

759.3611

LAXCAMSC3

SPRINTLD/2W

222

16:00

480

502

2.58%

16986

471.8333

IRVCA

MCI/2W

185

9:30

240

262

4.17%

8519

236.6389

LNBHCAXKCM2

GRDNCA0386T/2W

113

16:30

384

410

3.50%

13715

380.9722

 



[1] Please note that CROME had been in use for over four years at the customer site, allowing better resource management, detailed tuning and thus superior service even in a growing network infrastructure.   The same report will compete in under four (3.5) seconds across twenty-five (25) Nortel DMS switches with 2,726 trunk groups.  Note, Quantum constantly optimizes and improves the performance of the CROME PM suite thus you should expect higher performance numbers than these examples.