Background:
One year ago 3rd biggest telecommunication company sprint laid off 6000 employees and outsourced most of its network monitoring and management activities to off shore to reduce expenditure as a step towards globalization of the market economy .Also, hired Ericsson Services to provide network support in the local markets . Ericsson Services established centers in different parts of India to support these network management services for Sprint by hiring engineers on extremely low wages in those third world countries.
I have a project idea to consolidate network performance, outages and dashboard reports on one system as all in one solution by automating the process which traditionally handled by network engineers in different markets . By utilizing this idea wireless companies can save millions of dollars and ensure security of the users by using internal systems for the network footprints.
Moreover, it can save revenue, improve employees’ efficiency and certainly save jobs which have been outsourced in last one decade.
These reports can be run on central database in four different levels.
1: break fix activities
2: performance optimization activities.
3: executive level daily dashboard report
4: customer related logs, billing , call details etc

Break fix activites for technicians:
1: Running alarm log commands in every geographical area after every 5 minutes to collect outage reports and push the logs into a database with outage severity level and issue ID.
These reports can be collected on Vendor servers (mostly UNIX servers provided by Lucent to support their equipments called cell sites).

Complete Cell/Server outage or backhaul T1 outages as highest priority One or two components less service affecting as Moderate priority Power level difference with least priority.

Assign the issue id to available technician sending email with detail of the issue as well as address and access related information.
Provide interface to technicians to check assign jobs under their names and close issue after resolve.

Performance optimization level:

1: collect network log files from vendor servers for hourly performance of the sites, calls and key performance matrixes.

2: parse the logs into databases on Cell level hourly activity (these KPIs can be in 100’s).
3: provide interface for the users for the real time reports, with all required formulas to check degradation or improvement.
4: using performance reports draw maps to resolve any degradation.

Executive level daily dashboard report

Creating scripts to run reports using database for the previous day performance.
1: market level daily hourly reports
2: trending for the last 30 days to compare performance.

customer related logs, billing , call details etc
Customer billing can be done using the call logs from the database.
These features needs further support from billing department.
Since main database can provide call logs but not the customer profile or billing contract specification .

Cost analysis:

1: 2 Server cost 45000X2= $90,000 on time cost

The Oracle SPARC Enterprise M5000 midrange server delivers high scalability, mission-critical RAS, and advanced virtualization in a compact 10RU space. Oracle's SPARC/Solaris binary compatibility guarantee and the ability to mix several generations of processors provides excellent investment protection.

10 work stations: 2000 X 10= $ 20,000

2: developer cost:

4 programmers, to write scripting for reports, (Shell scripting, PHP scripting, perl scripting, excel Macros , web interface etc) 55,000,4=$ 220,000 per year

4: 5 network engineers to monitor the system 55000X 4= $ 220,000

3: Software cost (vortex to map performance report) $1000 per year. Using LAMP environment can safe money by using free software such as Linux, Apache web server,Mysql and Perl or PHP.

Miscellaneous Expenditure: $55,000

Total cost: 401,000 per year

Saving by elimination networking management group 55,000 X 25=1375,000 -401,000= $ 974, 000

Companies can save $ 974, 000 per year by implementing the automation process and discourage outsourcing.

StakeHolders:

1: Network management groups(Technicians, engineers and executives,)

2: Billing department groups

3: System users as cellular customers

4: System builders

5: Hackers (competitors and habitual hackers)

Project manager:

I have 10 years experience in developing network system management reports. Most recently, developed performance web and excel application for daily EVDO cell site performance for NY/NJ market. Reports were so appreciated in North East that I was asked to include 11 other markets in the database spanning 15 stats.

High level solution:

Database:

I am proposing Mysql database per physical location in the area

Tables:

Cell ID as primary Key

500-1000 performance matrixes.

Cloud computing can be utilized as a backup for the database since it available at very low cost in these days.

Design:

Databases per physical locations such as Brookly1, Brooklyn 2, Manhattan1 etc

Table for outages.

1: Cell id

2: Component OOS

3: severity level

4: issue ID

5: technician name

6: open or closed

Performance table:

Cell id

Blocks

Drops

Minute of usage

.........