Research On Radio Signal Strength Sensing With Pervasive Devices Within2.4GHz ISM Band

As an important technology for communicating, different wirelesscommunication technologies are more and more likely to work together.For example, GSM, GPRS, EDGE, WiMax, WiFi, Bluetooth, and ZigBeecoexist in many circumstances. Nowadays, different kinds of commonlyused wireless technologies work on the same2.4GHz ISM band (such asZigBee, WiFi and Bluetooth), interferences may occur among them.Although collisions decrease the performance of the wireless networks,coexistence and cooperation of different wireless networks introducescapacity to accomplish some specified tasks. This paper presents a systemthat enables other wireless devices to help sense signal strength in realtime. The design of this system is motivated by the observation that aheterogeneous radio can sense frame transmissions from other radio evenit cannot decode them. A radio passively listens on the wireless channeland estimates the RSS of a given client at its location. Based on this idea,we propose two systems.As the first system, we try to estimate the signal strength of anotherheterogeneous wireless radio. To tackle this issue, we faces three uniquechallenges. First, multiple radios may transmit frames concurrently andframe collisions may happen. Second, because of the diversity hardwarecapacity of wireless devices (for example, ZigBee devices have severeresource constraints), our approach should not assume that wireless radiocan sample the channel at arbitrarily high frequency and hence some frame transmissions may not be sampled. Third, distributedheterogeneous nodes are usually not time synchronized and on low-costradio (such as ZigBee) the time measured by local clock is inaccurate.To address these challenges, we propose a novel gateway-assistedapproach for estimating RSS values of a given wireless radio atheterogeneous nodes. A new light-weight algorithm is designed foridentifying the RSS values corresponding to the given radio. The centralidea is to search the subsequence of RSS samples on heterogeneous radiofor a digest sequence of the given radio. An optimization technique isproposed to address issues of time drift and time asynchronization. Wehave implemented a prototype under WiFi and ZigBee radios as example.More radios such as Bluetooth can be easily added in. Our extensiveexperiments on the testbed show that we can achieve low estimation error(~10%error rate), short delay (~8seconds) and small computationoverhead.As the second system, we used the idea of signal sensing todistinguish packet loss. Packet loss may due to different issues (weak link,collision, etc). Different packet losses need different mechanisms to dealwith. However, most currently used technologies do not distinguish them.Using this system, a ZigBee sensor network can diagnose the packet lossand adopt relative coping mechanisms.