FAM: A frame aggregation based method to infer the load level in IEEE 802.11 networks

In many environments, connected devices are exposed to and must choose between multiple Wi-Fi networks. However, the procedure for selecting an access point is still based on simple criteria that consider the device to be unique in the network. In particular, the network load is not taken into account even though it is a key parameter for the quality of service and experience. In this paper, we investigate how an unmodified vanilla device could estimate the load of a network in the user space with no interventions from the access points. In this regard, we propose a novel and practical method, FAM (Frame Aggregation based Method). It leverages the frame aggregation mechanism introduced in recent IEEE 802.11 amendments to estimate the network load through its channel busy time fraction. FAM combines an active probing technique to measure the actual packet aggregation and Markovian models that provide the expected rate as a function of the volume and nature of the traffic on the network. We validate the effectiveness of FAM against both ns-3 simulations and test-bed experiments under several scenarios. Results show that our method FAM is able to infer the network load with a granularity based on six different levels of network loads for the considered scenarios.

Automated summary

This paper investigates how an unmodified device can estimate network load in the user space without intervention from access points. A novel method called FAM is proposed, which leverages the frame aggregation mechanism to estimate network load and combines active probing technique and Markovian models for measurement and prediction. The effectiveness of FAM is validated through simulations and experiments.