Resource Allocation for Time-Variant Channels in the Nano-communication Networks

In this paper, regarding the limited energy resource of nanomachines, we propose a dynamic resource allocation scheme including frequency selection and power control for data transmission in nano-communication networks. Firstly, in view of the properties of THz communications, notably the sensitivities to the frequency and distance on the communication path, we employ Brownian motion to describe the mobility of nanomachine and build a time-variant channel model. Then, based on the channel model and energy resource of nanomachine, we develop a stochastic optimization to maximize the channel capacity. By virtue of the Lyapunov optimization theory, the problem of stochastic optimization is converted into a series of non-convex problems for the timeslots. The non-convex problem for each timeslot is further solved via the continuity relaxation and successive approximation theories. Moreover, we obtain a dynamic capacity sub-optimal resource allocation (CoRA) algorithm to accommodate the network only according to the current energy resource information and channel state information. Simulation results validate the theoretical analysis of our scheme.