Since 1999 IEEE 802.11 has become a dominating wireless technology for providing WLAN in both public and private places. The protocol has evolved with time and the current version of protocol allows the maximum data rate up to 6.76 Gbps. Providing high data rate makes it capable of supporting applications demanding high bandwidth and less delay. The protocol is built on carrier sense multiple access with collision avoidance (CSMA/CA) principle to minimize the collisions. In real-time networks as the traffic increases, collisions take place between different flows leading to delay and wastage of bandwidth. In literature, various works are dedicated to predicting the collision probability to estimate available bandwidth of the network. One of the popular approaches “ABE” (Sarr et al. in IEEE Trans Mob Comput 7(10):1228–1241, 2008) focused on this issue and calculates collision probability of “Hello Packets” and interpolating it to get the collision probability of data packets to estimate the available bandwidth. The current work proposes to use an analytical approach to predict the collision probability without any intrusiveness and estimate the available bandwidth for accurate admission control. Most of the literature that uses mathematical approach assumed homogeneous conditions. The current paper takes into consideration heterogeneous conditions to resemble real networks. The paper proposes amalgamation of passive approach and analytical approach, yielding more accurate results. The work is verified through extensive simulations to validate our proposed model. Since the model is based on analytical approach, major benefits are: (a) there is no need to send packets for estimating the collisions, hence scaling down the overheads; (b) One can change the parameters and predict the result set for any given conditions; (c) the method is cost-effective and non-intrusive in nature.

自1999年以来，IEEE 802.11已成为在公共场所和私有场所提供WLAN的主导无线技术。该协议随着时间而发展，当前版本的协议允许最大数据速率高达6.76 Gbps。提供高数据速率使其能够支持要求高带宽和较少延迟的应用。该协议建立在具有冲突避免（CSMA / CA）原理的载波侦听多路访问上，以最大程度地减少冲突。在实时网络中，随着流量的增加，不同流之间会发生冲突，从而导致带宽的延迟和浪费。在文献中，各种工作致力于预测冲突概率以估计网络的可用带宽。一种流行的方法“ ABE”（Sarr等人在IEEE Trans Mob Comput 7（10）：1228–1241中，你好数据包并对其进行插值以获得数据包的冲突概率，以估算可用带宽。当前的工作建议使用一种分析方法来预测碰撞概率而没有任何干扰，并估计可用带宽以进行精确的准入控制。大多数使用数学方法的文献都假设均质条件。当前的论文考虑了类似于真实网络的异构条件。本文提出将被动方法与分析方法结合起来，以产生更准确的结果。通过广泛的仿真验证了这项工作，以验证我们提出的模型。由于该模型基于分析方法，因此主要优点是：（a）无需发送数据包来估计冲突，从而减少了开销；（b）可以改变参数并预测任何给定条件的结果集；（c）该方法具有成本效益，并且是非侵入性的。