A Decentralize Algorithm for Perturbation Minimization in 5G D2D Communication

Abstract

In 5G device to device (D2D) communication, two users residing in close proximity can directly communicate among themselves without the need of a base station. Such a communicating pair of users forms a D2D link which operates on a common channel. Suppose $Y$ dedicated channels are available to support the D2D communications. Since channels are limited resources, we may have to reuse the same channel to multiple links in order to enhance the capacity. But if two links are interfering to each other, they cannot use the same channel to avoid co-channel interference. The interference constraints at a particular time instant $t$ can be modeled in terms of a graph $g(t)$ where each vertex represents a link and two vertices have an edge between them if their corresponding links are interfering to each other. At any time instant, a vertex must be assigned with a channel, otherwise the corresponding user pair can not communicate. Due to the movement of the users, a color/channel assigned at time $t-1$ may not be a valid coloring at time $t$. Hence we may have to recolor some vertices to satisfy the interference constraints. Our objective is to minimize the total number of such recoloring/perturbations $A(t)$ at time $t$. To minimize $A(t)$, we propose a decentralize differential coloring (DDC) algorithm and calculate the asymptotic bound on $A(t)$ produced by DDC. We also compare the value of $A(t)$ produced by DDC with other existing approaches. Results show that DDC produces less perturbations than the existing approaches. Theoretical results have also been validated through simulations.