Telecom operators need automation in wireless infrastructure

Wireless operators have grown networks to a point past the level of complexity that can be managed and optimized manually.  Today, wireless networks have up to 6-layers at the macro level with additional micro levels inserted in key areas where network capacity is an issue.  In order to manage the complexity of interoperability handoffs and signaling necessary to provide the highest call quality, operators need automation.

Wireless operators have grown networks to a point past the level of complexity that can be managed and optimized manually.  Today, wireless networks have up to 6-layers at the macro level with additional micro levels inserted in key areas where network capacity is an issue.  In order to manage the complexity of interoperability handoffs and signaling necessary to provide the highest call quality, operators need automation.

Today’s wireless networks capture every step in a mobile call, from origination through the MSC (Mobile Switching Center), to the termination of the call.  This data is captured and stored throughout the key infrastructure in the network, like MSC, RNC (Radio Network Control) logs, OSS (Operations Support System), HLR/VLR (Home/Visitor Locate Registers) and various other nodes.  This subscriber and network based intelligence is a gold mind of data that tells the carrier exactly what occurred throughout the life cycle of a mobile call.

One automated method to gather this data, process it and paint a picture of the event uses Chaos theory.  With all the adjustments, processed by different nodes, handovers, power changes, signal measurements an more…it is amazing that calls are even completed.  However, processing the collected data through a series of Chaos theory algorithms, the data can be organized and compartmentalized into its respective categories, showing the orderly progression of the call and painting a picture of what occurred.  These pictures can be overlaid on geographic maps, metropolitan roads and streets giving the operators keen insight into where network issues exist and where they do not.

Unlike test drives that cover only finite periods of time and specific routes, the data captured on the network utilizing this automated method can be processed for any time and location in the network.  Using this captured intelligence to develop models around trending for capacity management, optimization in difficult areas and/or buildings, as well as performance enhancements, solutions can be tested prior to implementation into a live network.  Automated tools can even suggest network modifications to improve performance.

The example below shows data collected from two RNC’s in a wireless network.  RNC’s are devices that manage macro cell sites.

The overall Inter-Radio Access Technologies Hand Over (Inter-RAT HO) per call for these two RNC’s is 0.28; meaning about one in four calls are directed to the GSM network.  LRNC002 has 40% of calls directed to GSM, equating to 2.5 calls directed to the GSM network.  Overall a 5% Inter-RAT HO failure is experienced, shown in Table A.

 

 

Table A

RNC

Voice Call Attempt

Voice Soft Handover

Soft-Handover per Call Attempt

Inter-RAT Hard Handover Attempt

Inter-RAT Hard Handover Failure

Inter-RAT Handover Failure Rate

Inter-RAT Handover per Call Attempt

ARNC002

11,600

42,524

3.67

1,432

28

1.96%

0.12

LRNC002

15,585

116,117

7.45

6,295

366

5.81%

0.40

Total

27,185

158,641

5.84

7,727

394

5.1%

0.28

 

 

 

The system was asked to make recommendations to improve Inter-RAT HO in the two RNC’s analyzed. The data shown in Table B are the results based on the implementation of the system’s automated suggestions.  The Inter-RAT reselection is now 1.06 per call, resulting in a 3G cell update success rate of 97.41% and an Inter-RAT cell reselection success rate of 99.40%

 

 

 

Table B

 

RNC Voice call attempt Cell Update Attempt
(3G->3G)
Cell Update Success (3G->3G) Cell Update Success Rate  (3G->3G) Cell Update (3G->3G)
per Call
Inter-RAT Cell Reselection Attempt Inter-RAT Cell Reselection Success Inter-RAT Cell Reselection Success Rate      (3G side) Inter-RAT Cell Reselection Attempt per Call
ARNC002 11600 1745 1699 97.36% 0.15 8697 8630 99.23% 0.75
LRNC002 15585 7082 6899 97.42% 0.45 20017 19911 99.47% 1.28
Total 27185 8827 8598 97.41% 0.32 28714 28541 99.40% 1.06
 
 

Clearly, technology is available with proven algorithms to make solid suggestions, based on real network intelligence.  With full end-to-end network intelligence, operators can fine-tune their networks to normalize capacity, improve signal management and test solutions before implementation.  This gives operators the ability to leverage subscriber and network based intelligence to create an effective and efficient model for wireless network performance management.

The automation of the data to create models and suggest solutions will assist operators in managing CAPEX and OPEX around the complex multilayered networks that customers depend on everyday for their data and voice communications.

Eric Moore is the Chief Operating and Technology Officer at Axis Teknologies, LLC, a wireless engineering and services company with expertise in 4G CORE and RAN engineering, services and optimization.  Moore can be reached at emoore@axisteknologies.com or follow Axis Teknologies on Twitter @axisteknologies.