Statistical Transition Map for WLAN Roaming Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current enterprise wireless local area networks (WLANs) face challenges in managing roaming for delay-sensitive multimedia applications due to high mobility of user devices, impacting network performance and user experience.
Innovation Solution
A statistical transition map is generated based on mobile device mobility history data, using received signal strength and location trace information converted into natural language pseudo-location word labels, allowing for predictive roaming and anomaly detection by computing the probability of next locations based on current positions and historical patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional roaming management is used in enterprise WLANs, then network deployment is simpler, but roaming performance and user experience deteriorate due to high device mobility and delay-sensitive multimedia applications
Solution Approach 1:
The system performs preliminary actions by collecting mobility history data and generating statistical transition maps in advance, before actual roaming events occur. This allows the network to predict future device locations and prepare optimal handover decisions proactively, improving roaming performance while managing complexity through pre-computed models
Solution Approach 2:
The patent creates a virtual copy of the physical environment by generating a statistical transition map that represents mobility patterns without exposing absolute physical locations. This abstracted model enables intelligent roaming decisions while simplifying network management by working with relative position data rather than complex physical network configurations
2Measurement precision
If absolute physical locations are tracked for predictive roaming, then roaming accuracy improves, but user privacy and security worsen due to exposure of sensitive location information
Solution Approach 1:
The system applies local quality by using relative location information and pseudo-location word labels that are meaningful for predicting next positions but do not reveal absolute physical locations. Each location is represented locally within the statistical model context rather than as global coordinates, maintaining prediction accuracy while protecting privacy
Solution Approach 2:
The patent introduces an intermediary layer - the statistical transition map with pseudo-location labels - that mediates between physical location tracking and privacy protection. This intermediary representation enables accurate roaming predictions by capturing mobility patterns without directly exposing sensitive absolute location data to the system or users
Data Source
AI summary
A statistical transition map is built based on mobile wireless device user mobility history data. This data is useful to assist various wireless local area network applications. Received signal strength and location trace information associated with movements of mobile wireless devices in a wireless network is collected. The received signal strength and location trace information is converted to a sequence of natural language pseudo-location word labels representing pseudo-locations of each mobile wireless device as each mobile wireless device moves about with respect to a plurality of wireless access point devices in the wireless network. A statistical transition map is generated for each mobile wireless device from the sequence of natural language pseudo-location word labels using a natural language model. A probability of a next pseudo-location for a particular mobile wireless device is computed based on its current location and its statistical transition map.