Performance modeling and optimization of multiple‐objective cross‐layer design in multi‐flow ad‐hoc networks

Abstract

journal/dac 1 of 25 wireless links. Most works in the existing literature have focussed on protocol implementation in a particular layer of the traditional networking stack. Therefore, changes in one layer do not require changes in other layers of the protocol stack. This strictly layered architecture leads to suboptimal design of wireless networks. This is because the implementation of one layer is completely decoupled from the other layers in the conventional network models. To deal with nontrivial challenges of wireless communication (eg, channel errors, varying network conditions, packet collisions) and to optimally utilize the limited resources (eg, energy and bandwidth) of the inherently resource-constrained wireless nodes, the concept of cross-layer information exchange is employed.Relative to their wired counterparts, the high-speed real-time applications such as multimedia messaging, remote sensing, audio and video conferencing exhibit several unique characteristics such as stringent delay constraints, demand for high data-rates, etc. The design of ad-hoc wireless networks to implement these multimedia services frequently involves the optimization of various desirable design parameters while satisfying the physical resource constraints of the network. As a result, multiple-criterion optimization formulation has been widely adopted in literature for efficient design and operation of the networks.Throughput and latency are the fundamental and essential metrics of performance measurement in the network management layer to provide guaranteed quality of service (QoS) support in network design. Optimizing the overall throughput and the end-to-end delay of each session in the network is crucial for real-time applications as well as for data validation. Increasing network utilization for successful transmission of all packets and reliability of packet deliveries by considerably increasing the throughput is a significant issue in ad-hoc wireless networks. On the other hand, excessive delay can render most of the multimedia applications such as VoIP, high-throughput file transfers, distributed gaming, video-streaming services, etc. inoperative as they are delay sensitive. The end-to-end delay is composed of several smaller delays that collectively add together to compute the metric. For instance, delays due to contention, queuing, propagation, and transmission time in the network are usually included in the total delay parameter. Hence, the network delay parameter must be reduced to a minimum for real-time networks.Moreover, real-time applications in the context of ad-hoc wireless networks not only require reliability and timeliness constraints, but data must be routed through energy-efficient paths in the network to maximize the network lifetime. The energy-constrained operation of wireless nodes is attributed by the fact that each node is equipped with a non-rechargeable and irreplaceable battery as a power source. A node can act both as a data source and a router that relays data for other nodes. In addition, t...

Multi‐objective design of energy harvesting enabled wireless networks based on evolutionary genetic optimisation

“…Furthermore, our proposed optimisation model is compared with the earlier existing works [2, 14, 15, 16, 20 ] in the literature in terms of the following performance parameters: throughput, energy efficiency, and packet loss rate. These parameters to a certain level relate to each of the objectives that our optimisation procedure seeks to attain.…”

“…Also, the authors in [16 ] deployed a Matlab simulator for network lifetime and energy harvesting optimisation in the two‐ray path loss channel model under the network conditions of outage probability and spatial frequency reuse. Finally, our previous work in [20 ] applied the CVX software to jointly optimise the throughput, energy, and delay characteristics of the ad hoc wireless network through the cross‐layer cooperative Lagrangian and convex optimisation techniques.…”

“…This design model based on the mathematical integer linear programming considered dense traffic scenarios with minimal path length and aimed at mitigating the interference power and improving the packet loss rate and network throughput. In our previous work [20 ], the three weighted objective functions of increased throughput, attenuated delay, and energy‐efficient data transmission are jointly optimised through cross‐layer multi‐objective paradigm. The objective of minimising energy usage is enforced by maximising the residual battery energy available in each active node in the network, as opposed to the sustainable energy harvesting mechanism from the environment.…”

“…The objective of minimising energy usage is enforced by maximising the residual battery energy available in each active node in the network, as opposed to the sustainable energy harvesting mechanism from the environment. Moreover, the implementation of the developed optimisation model [20 ] is compared with the standard multi‐objective genetic algorithm technique in accomplishing the substantial performance gain. An iterative multi‐objective algorithm is proposed in [21 ] to explore the trade‐off between spectrum efficiency and harvested energy for multi‐cell MIMO networks subject to the transmission power restraints.…”

Adaptive immune‐inspired energy‐efficient and high coverage cross‐layer routing protocol for wireless sensor networks

“…The dissimilar layers interchange data in a non‐hierarchical method. To attain the necessities of QoS, the idea of the modular layers, optimising the design of network protocols, and dropping the overhead are disturbed by the numerous protocols following the cross‐layer model [10, 11].…”