Probe Server Parallelism Management for Wireless Traffic Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Monitoring and analyzing wireless traffic in cellular networks is computationally expensive due to the large volume of data, especially when real-time monitoring is required, which increases complexity and cost for service providers.
Innovation Solution
A two-layer architecture using passive Deep Packet Inspection (DPI) with probe and aggregator servers, employing optical splitters for data tapping, load balancing, and hashing to manage parallelism based on available resources, optimizing stream processing on Non-Uniform Memory Access (NUMA) machines and correlating end-to-end measurements through control plane monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If real-time monitoring of wireless traffic is implemented, then the speed of traffic analysis is improved, but the computational cost and device complexity increase significantly
Solution Approach 1:
The patent divides the traffic monitoring system into multiple probe servers that process different streams of traffic data in parallel. Each probe server handles a portion of the overall traffic load, enabling real-time processing without requiring a single complex system to handle all traffic simultaneously. This segmentation reduces the computational burden on individual servers while maintaining real-time analysis capabilities.
Solution Approach 2:
The patent introduces the dimension of parallel processing by deploying multiple probe servers that operate simultaneously on different traffic streams. This transforms the problem from a single-threaded sequential processing model to a multi-threaded parallel model, effectively increasing processing speed without proportionally increasing the complexity of each individual server.
2Quantity of substance
If the volume of wireless traffic data to be monitored increases, then the quantity of information available for analysis is improved, but the computational cost increases
Solution Approach 1:
The patent implements a sampling approach where not all traffic data is processed with the same level of detail. Instead, the system processes a representative subset of traffic streams in full detail while using aggregated or summarized data for other streams. This partial processing approach provides sufficient analytical insight without the computational expense of analyzing every single data point at maximum detail.
Solution Approach 2:
The patent extracts only the most relevant features and metrics from the raw traffic data for detailed analysis, while other data is processed at a higher level of aggregation. This extraction of essential information allows the system to work with a reduced data set that retains the key analytical value while consuming less computational energy.
3Productivity
If the number of probe servers is increased to handle more traffic streams, then the productivity of traffic processing is improved, but the device complexity and cost increase
Solution Approach 1:
The patent designs probe servers with standardized, multi-functional capabilities that can handle various types of traffic streams using the same processing architecture. Each probe server is configured to perform identical functions across different traffic domains, allowing the system to scale by simply adding more units of the same standardized component rather than designing increasingly complex specialized systems.
Solution Approach 2:
The patent enables dynamic adjustment of processing parameters such as the number of parallel streams, sampling rates, and analysis depth based on available resources and traffic conditions. This allows the system to optimize productivity by changing operational parameters rather than always requiring additional hardware, thereby scaling processing capacity without linearly increasing system complexity.
Data Source
AI summary
A method, computer readable medium and apparatus for managing a degree of parallelism of streams are disclosed. For example, the method analyzes wireless traffic to determine a profile of the wireless traffic, determines an amount of available processing resources, and manages the degree of parallelism of streams based on the profile and the amount of available processing resources in a probe server.