This is a readme file for research data associated with 'Energy Harvesting Aided Device-to-Device Communication in the Over-Sailing Heterogeneous Two-Tier Downlink'

Abstract: Device-to-Device (D2D) communication and heterogeneous networks have been considered as promising techniques for alleviating the demand both for increased spectral resources and for additional infrastructure required for meeting the increased tele-traffic. 
For the sake of improving both the bandwidth efficiency and the network capacity of heterogeneous cellular networks constituted by multiple tiers, a direct D2D communication is arranged between a pair of nearby devices without involving the base station (BS), whilst reusing the cellular resources. We aim for maximising the sum-rate of the energy harvesting (EH) aided D2D links in a two-tier HetNet by superimposing their messages on the downlink resources of mobile users (MUs), which is achieved without unduly degrading MU's throughput. Specifically, our optimization problem relies on the objective function of maximising the D2D sum-rate based on the joint assignment of both the resource blocks (RBs) and of the transmission power for both the EH aided D2D links and for the MUs. This non-convex optimization problem, which is intractable in its original form, is then converted to a tractable convex problem, which is then analyzed by invoking the method of Lagrange multipliers of constrained optimization. As a result, an algorithmic solution defined as 'joint optimization of RB and power allocation (JORPA)' is proposed, which jointly allocates the RBs and power for the D2D links, whilst relying on the results of Lagrangian constrained optimization, when the base stations (BSs) of different tiers obey one of the following regimes: (a) orthogonal, (b) co-channel and (c) the proposed co-orthogonal channel deployments. We also propose low complexity heuristic methods for optimising the D2D transmit power while defining the D2D-MU matching heuristically and vice versa. The performance of both the JORPA algorithm as well as of the low-complexity heuristic algorithms is quantitatively analysed using our simulation results for different channel deployments relying on diverse network parameter settings. As expected, orthogonal deployment performs best, followed by  the co-orthogonal  and co-channel deployments. Moreover, the throughput experienced by the MUs in presence of D2D communication is guaranteed by our co-orthogonal scheme as well as orthogonal scheme, while co-channel suffers a marginal degradation when compared with throughput threshold. We also demonstrate that our 'equal power allocation (EPA)' heuristic method is capable of achieving 96% of the sum-rate achieved JORPA while other heuristic methods perform less well, implying that the optimization of the D2D-MU matching is indeed crucial for the system considered.

This excel file contains data for following figures :

Fig. 4: Convergence of of are algorithm as a sum-rate of D2D links with respect to the number of iterations. Here P = 1, D = 8, C = 10, B = 20dB, Rp = 4bps/Hz, and Rc = 8bps/Hz.

Fig. 5: Effect of the interference-threshold for D2D links from the PBS on the D2D sum-rate. Here P = 1, D = 8, C = 10, B = 20dB, Rp = 4bps/Hz, and Rc = 8bps/Hz.

Fig. 6: Impact of varying the number of D2D links on the D2D sum-rate. Here P = 1, C = 10, D = 8, I_Th = 40dBm, Rc = 8bps/Hz and Rp = 4bps/Hz.

Fig. 7: Impact of varying the number of Pico Base stations on the D2D sum-rate. Here D = 8, C = 10, I_Th = 40dBm, B = 20dB, Rp = 4bps/Hz and Rc = 8bps/Hz.

Fig. 8: Impact of the Quality of Service threshold of MU associated with MBS on the D2D sum-rate. Here P = 1, D = 8, C = 10 I_Th = 40dBm, Rp = 4bps/Hz and B = 20dB.

Fig. 9: Impact of the Quality of Service threshold of the MU associated with PBSs on the D2D sum-rate. Here P = 1, D = 8, C = 10, I_Th = 40dBm, Rc = 8bps/Hz and B = 20dB.

Fig. 10: Impact of varying the number of MUs on the D2D sum-rate. Here P = 1, D = 8, I_Th = 40dBm, B = 20dB, Rp = 4bps/Hz and Rc = 8bps/Hz.