Design and Development of Optimization Agent in Cross Layered Framework

A Cross Layered framework is an important concept in today’s world given the abundant usage of both single-path and multi path wireless network architectures. One of the important design issues in the development of a robust framework such as this is the design of an Optimization Agent or an OA. In recent days of wireless and wired ad-hoc networks, cross-layer design was brought about a few years back to explore attached optimization at different layers. In order to describe solutions in cross-layered design, the Open System Intercommunications model was employed. However, it is clear that no voice and reference mechanism exists to aid optimization, which could effectively halt effective adaptability and deployment of cross-layered solutions. In this study, we suggest some hypotheses regarding how to model and create cross-layer solutions using the OSI layered method. We use the aforementioned method to analyse and simulate a particular type of cross-layered solution, namely energy-aware routing protocols. We use a layered approach to examine two proposals that are accessible in the literature. The applied strategy leads to the creation of


I. INTRODUCTION
Cross-layer design and optimization may be a novel technique that has become commonplace in recent years and enhances network performance in both wired and wireless networks. The primary concept behind cross-layer design is to optimise data control and exchange across two or more layers in order to achieve considerable performance gains by leveraging interactions between different protocol layers.
In this research, we offer a cross-layer design and optimization framework, as well as the idea of using an optimization agent to provide data interchange and control between various protocol layers in wireless sensor networks to improve performance.
The approach is to investigate the effects of the wireless channel on the physical layer performance of a small-scale wireless sensor network (WSN) in order to get insights that will be used to design and build the optimization agent within the proposed crosslayer framework.
For a group of networked wireless sensors, performance measurements were taken to examine the effects of interference and transmission range.
The development of mobile ad-hoc wireless networks (MANETs) is inspired by a wide range of enticing applications that occur in dynamic circumstances where the construction of a communication infrastructure would be difficult to achieve using a different type of network.
Many of these difficult application cases impose significant network design constraints. In general, network nodes must be small, adaptable to a wide range of environments (e.g. body networks), and deployable in large numbers (i.e. dense networks); cheap, to ensure dense network deployment is costeffective; battery-powered, to ensure autonomy and freedom of movement while simultaneously forcing the energy consumption issue to become the most important concern in order to maximise efficiency.
Communication resources may be limited as a result of these differences. Low bit rates, in particular, must be used to keep the transceiver simple while maintaining the specified bit error rate (BER).
Furthermore, as the number of active nodes in the network grows, the supply of the channel diminishes.
A high number of nodes, on the other hand, promotes high connection among them. High connectivity is typically used to reduce energy by employing a multihop strategy, as well as to increase dependability by utilising redundancy in data collecting and message delivery.
The requirement of self-configuring, maintaining, and adapting the network to achieve the desired performance drives communication protocol design in this environment. The introduction of new functions that make efficient use of available resources makes it necessary to advance the way networks are designed.
Traditionally, networks have been created using the OSI model [3], in which nodes are logically split into layers and placed in an extremely stackable manner.
Interaction between adjacent layers is enabled by defining interfaces between them. The network design is broken into an extremely large number of discrete and easier design activities once each interface has been correctly described by exposing services but hiding their implementation.
Cross-layer design could be a new way to build and operate networks, claiming a greater level of interaction between layers to develop solutions that adapt to changing conditions and application requirements [2]. The goal is to improve performance in a resource-constrained environment.
Despite the extensive literature on cross-layer optimization methods, a good definition of this methodology is still lacking. Furthermore, the functional description of cross-layer solutions is done in an unstructured way, despite the fact that the need for a reference model has been recognised in.
We analyze this with the following approach: We develop a correct API for each layer to facilitate cross-layer interaction and we locate various elements of the analysed algorithms within the reference layers.
Our strategy is based on two main points: 1) It demonstrates that the OSI model's inapplicability to cross-layer design must be properly justified before claiming the need for a replacement reference model. The wireless channel, unlike wire line networks, has various distinct properties that must be taken into account while developing wireless networks [1]. As a result, we now understand the significance of an agent-based optimization system, or an optimization agent within multi-layered frameworks. In the next section, we'll look at two different designs for agent-  As illustrated in Figure 1, we present a generic framework for our study of cross-layer optimization for WSNs.
The architecture consists of a proposed optimization agent (OA), which facilitates interactions between various protocol layers by serving as a core repository or database where essential information such as node positive identification, hop count, energy state, link status, and so on is temporarily stored and used as side information for other layers across the protocol stack.
Information can only be shared directly across two adjacent layers in an extremely sequential manner, which differs from the layered model approach.

Intra-layer interactions (between adjacent layers) and
inter-layer interactions (across two or more adjacent layers) will be classified, and these interactions will be either bottom up or top down [7].  To achieve this purpose, a general cross-layer framework was presented, as well as the idea of using the OA to improve and optimise the performance of a wireless system. It would be required to understand the features and performance of each protocol layer, as well as how they interact with one another, before creating the OA and developing the cross-layer