ElM: Efficient online Interference Measurement in Wireless Sensor Networks

In wireless sensor networks, knowing real signal interferences (received signal strength or RSS) from other sensors to cared sensor is critically important for protocol design and for many applications. However, the real interferences generally differ much from those calculated by theoretical or empiri­ cal models, because these models cannot capture the dynamic impacts of the environments. This paper presents ElM, an efficient method for sensors to online measure their interference vectors, where the interference vector of a sensor indicates the real interferences from other sensors to it when other sensors are transmitting packets. ElM conducts interference vector measurement passively during the natural working process of the sensor network. Because a sensor may be interfered by quite a few neighbors and the environments may change over time, the efficiency i.e. latency for real-time interference vector estimation is a very challenging issue. ElM exploits three facts to improve the efficiency of interference vector calculation. First, by exploiting the additivity property of the received signal strength, a linear interference vector reconstruction model is developed. Secondly, ElM exploits not only the received and overheard packets from other sensors, but also the collided packets to construct the observation matrix of the linear model. Third, compressive sensing technologies are developed to recover the interference vector even when the observation matrix is partly determined. These methods help sensors much reduce the latency for calculating real-time interference vectors. Extensive simulations were carried out to verify the effectiveness and efficiency of the proposed methods.