Distributed Clustering and Spectrum-Based Proximity Device Discovery in a Wireless Network

Abstract

Direct short-range device-to-device (D2D) conununication enjoys many advantages, such as high transmission rates, low propagation delays, low transmit power and interference. In order to realize the advantages, a device or user equipment participating in the D2D communications have to determine that the devices discovered are in a proximity area and are suitable to establish links with them. Hence, an accurate and efficient clustering scheme, which can organize the devices nearby into multiple D2D communication groups is necessary. Against this background, in this paper we propose a self-clustering scheme which requires no assistance from Cellular Network. The philosophy of our design is to group the devices with similar sensed spectrum profiles. In order to facilitate this self-clustering scheme, we adopt Reed-Solomon (RS) code into our proposed design. More specifically, in our design, a D2D communication group is formed by the D2D devices whose sensed spectrum vectors are mapping to the same valid RS codeword which is broadcast by the form of frequency-hopping (141) pattern. Additionally, the collision event could happen when multiple D2D devices transmit signals to discovery their group members. Hence, we design the FH probing sequence based on a randomly cyclic-shifted non-binary RS codeword to eliminate the possible collision events and interference exited in clustering procedure. In this paper, we provide our detail sequence design and the self-clustering procedure. In the simulation results, we show that the proposed self-clustering is efficient and effective.