No abstract

Mobile edge computing (MEC) has shown tremendous potential as a means for computationally intensive mobile applications by partially or entirely offloading computations to a nearby server to minimize the energy consumption of user equipment (UE). However, the task of selecting an optimal set of components to offload considering the amount of data transfer as well as the latency in communication is a complex problem. In this paper, we propose a novel energy-efficient deep learning based offloading scheme (EEDOS) to train a deep learning based smart decision-making algorithm that selects an optimal set of application components based on remaining energy of UEs, energy consumption by application components, network conditions, computational load, amount of data transfer, and delays in communication. We formulate the cost function involving all aforementioned factors, obtain the cost for all possible combinations of component offloading policies, select the optimal policies over an exhaustive dataset, and train a deep learning network as an alternative for the extensive computations involved. Simulation results show that our proposed model is promising in terms of accuracy and energy consumption of UEs. INDEX TERMS Computational offloading, deep learning, energy efficient offloading, mobile edge computing, user equipment.

No abstract

No abstract

In homogeneous cellular networks, fractional power control (FPC) is employed to partially compensate the path-loss and, hence, improve uplink (U L) signal-to-interference ratio (SIR). However, this scheme is less effective in heterogeneous cellular networks (HetNets) because: (i) except the typical user, all other users with variable U L transmit power (UTP) act as interferers, (ii) FPC leads to high UTP by edge users and, hence, more interference, and (iii) small base stations (SBSs)' densification further increases network interferences. Leveraging FPC in HetNets, we propose nonuniform SBS deployment (NU-SBS D) to reduce interference and, thus, increase network performance. According to our NU-SBS D model, SBS deployment (SBS D) near macro base station (MBS) is avoided, whereas MBS coverage edge area is enriched with ultra-dense SBS D. NU-SBS D model leads to: (i) better SIR reception of MBS coverage edge users, (ii) fewer SBS D requirement, and (iii) better SBS coverage in the MBS coverage edge area. Moreover, to make a model more proactive, we also consider reverse frequency allocation (RFA) to further abate both U L and downlink (D L) interferences. The coverage probability expressions are derived for both uniform SBS deployment (U-SBS D) and NU-SBS D while using RFA and FPC. Through simulation and numerical results, we characterize coverage probability for different values of SIR threshold, path loss compensation factor, SBS density, users density, and the distance between the typical user and the associated base station. The proposed NU-SBS D model along with RFA leads to reduced network interference as compared with U-SBS D and, thus, leverages FPC in HetNets.

No abstract

Vehicular ad hoc networks play a pivotal role in the enrichment of transportation systems by making them intelligent and capable of avoiding road accidents. For transmission of warning messages, direction-based greedy protocols select the next hop based on the current location of relay nodes towards the destination node, which is an efficient approach for uni-directional traffic. However, such protocols experience performance degradation by neglecting the movement directions of nodes in bi-directional traffic where topological changes occur dynamically. This paper pioneers the use of movement direction and relative positions of source and destination nodes to cater to the dynamic nature of bi-directional highway environments for efficient and robust routing of warning messages. A novel routing protocol, namely, Direction Aware Best Forwarder Selection (DABFS), is presented in this paper. DABFS takes into account directions and relative positions of nodes, besides the distance parameter, to determine a node's movement direction using Hamming distance and forwards warning messages through neighbor and best route discovery. Analytical and simulation results demonstrate that DABFS offers improved throughput and reduced packet loss rate and end-to-end delay, as compared with eminent routing protocols.

In heterogeneous cellular networks (HetNets), small base stations (SBSs) are overlaid in the coverage region of a macro base station (MBS) to improve coverage and spectral efficiency. However, the performance of HetNets is significantly degraded by inter-cell interference (ICI) due to aggressive frequency reuse and multi-tier deployment. Besides ICI, the uplink (UL) communications of MBS edgeusers (M-EUs) are prone to jammers' interference (JI) due to wide-band jammers (WBJs). With sufficient knowledge of network parameters, such as frequency bands and transmit powers, WBJs inject JI in the UL communications band to affect legitimate communications by degrading UL signal-to-interference ratio (SIR). Such distributed denial-of-service (DDoS) attacks normally target organizations, shopping malls, or public gatherings by clustering around them. As a countermeasure, we use decoupled association (DeCA) for the M-EUs, as opposed to the coupled association (CA), to improve UL SIR. Additionally, we use proactive interference management scheme, known as reverse frequency allocation (RFA), along with DeCA to resist both ICI and JI. The results show that WBJs cluster effectively degrades the legitimate UL communications of the target. The results also demonstrate that the network performance degrades significantly by increasing jammers' density and transmit power. Furthermore, DeCA with RFA leads to improved network performance due to effective ICI and JI mitigation. INDEX TERMS Coverage probability, denial-of-service, decoupled association, heterogeneous cellular networks, matern cluster process, poisson point process, reverse frequency allocation, wide-band jammers.