Future SON Tech: MIT Test Wireless Network Performance Optimization Using Sensor Hints

Posted · Add Comment

Here is clear case of Extra Sensory Perception (ESP.) Researchers Lenin Ravindranath, Calvin Newport, Hari Balakrishnan and Samuel Madden at MIT Computer Science and Artificial Intelligence Laboratory recently put together some prototype network improvements to the test over WiFi. The concept was to integrate local information, such as from sensors, to improve network performance. They used the GPS, motion and position sensors to create ‘hints’ for their improvement algorithms and hints are communicated using a specific protocol they developed for the testing. There were 4 algorithms including (a) Hint aided Rate Adaption (LTE has adaptive modulation so this is easy to add) (b) Optimization of WiFI Access Point (AP) Association (c) Mobile Topology Maintenance and (d) Path Selection in Vehicular mesh networks. The result of the testing was a roughly 50% net throughput increase in most situations and some handover reductions in others. To me this is very usable research and easily implemented, with good fit, into future implementations of 3GPP LTE SON. Check it out…

Link: MIT Computer Science and Artificial Intelligence Laboratory, MIT News


| | | Next → |