Variable rate m-QAM assisted Packet Size Optimization for Cognitive Radio and MIMO Cognitive Radio based Sensor Networks

This paper proposes a novel system model where the concept of Cognitive Radio is introduced alongside conventional sensor network framework operating within the unlicensed ISM frequency band. The sensor nodes with channel sensing capacity are considered to be cognitive secondary user (SU) and the one without channel sensing feature are designated to be as the primary user (PU). Under the premise of our proposed system model, a joint optimization framework is formulated to determine the physical layer (PHY) optimal packet size(OPS) for the cognitve radio based sensor network (CRSN) model in case of a linear equidistant point to point multihop network topology. It is observed when fixed rate modulation scheme like FSK is used in the context of CRSN, it fails to provide the OPS value under unfavourable network conditions like interference and overall transmission timing constraints. To overcome this problem we apply variable rate m-QAM based modulation scheme to the CRSN framework. Furthermore our concept is extended to a Multiple Input Mutiple Output (MIMO) supported Cognitive Radio based Sensor Network (MIMO+CRSN) which provides an OPS value under more unfavourable network conditions. At signal bandwidth of 20 KHz and number of available ISM channels equals to 30, it is observed that fixed rate FSK based point to point CRSN architecture could support 3 cognitive secondary users in terms of OPS determination. Whereas the proposed point to point variable rate m-QAM based framework provides an OPS value for 10 users with an increase in value by 20%. For the extended MIMO+CRSN framework, it could support 10 users with an increase in OPS value by 50% at 7 users under primary transmit power of 10dBm and more severe average primary user interference time constraint.