Fair airtime allocation for content dissemination in WiFi-direct-based mobile social networks

Abstract

The vast penetration of smart mobile devices provides a unique opportunity to make mobile social networking pervasive by leveraging the feature of short-range wireless communication technologies (e.g. WiFi Direct). In this paper, we study local content dissemination in WiFi-Direct-based mobile social networks (MSNs). We propose a simple GO-coordinated dissemination strategy, as WiFi Direct does not originally support content dissemination. Due to mobility and the short transmission range, the duration of nodes in contact tends to be limited and consequently they compete for the limited airtime to disseminate their own data. Therefore, fair allocation of the limited airtime among the nodes is required. We focus on fairness in content dissemination rate, which is a key application-layer metric, rather than fairness in throughput or airtime and formulate the allocation problem as a generalized Nash bargaining game wherein the nodes bargain for a share of the limited airtime. The game is proved to have a unique optimal solution, and an algorithm with low complexity is designed to find the optimal solution. Furthermore, we propose a detailed scheduling approach to implement the optimal solution. We also present numerical results to evaluate the Nash bargaining based allocation and scheduling.

Joint head selection and airtime allocation for data dissemination in mobile social networks

“…This social property has not been considered in previous studies on airtime allocation in WLANs, cellular networks or mobile ad hoc networks (MANETs). In our earlier work [36], we propose a fair airtime allocation scheme for group data dissemination based on NBS. However, this scheme assumes the group head is already selected, and does not characterize user specifics including data dissemination need, preference on data, energy cost, etc.…”

Optimal Group Formation in Dense Wi-Fi Direct Networks for Content Distribution

“…The study involves random GO selection based on the node's requirement to publish contents. A slightly different and simpler method for GO selection is proposed in [36] that is based on the amount of data load for each device. The device which has the largest data to transmit is selected as a GO.…”

“…The works in [13] and [40] propose back-up GO selection and redundant group formation respectively in scenarios where the P2P group is frequently terminated due to network dynamics. The works stated earlier [1], [11]- [13], [34], [36] requires the modifications of standard Wi-Fi Direct functions to improve the performance. The work in [25] further supports our idea of modifying the Probe Request and Probe Response frames, in their work to implement Wi-Fi Direct based multi-hop emergency communication network.…”

“…To validate the benefit of the proposed scheme in terms of throughput gain, the proposed scheme is compared with other methods discussed earlier in Section IV, i.e. standard group formation [2], RSSI based scheme [34] and load-based scheme [36]. To perform the comparison, multiple simulations are performed using random STAs' placement.…”

“…The proposed scheme improves throughput gain by 18% and 54% as compared to RSSI-based and load-based schemes. The reason for higher throughput gains using the proposed scheme is that it considers the actual link rates available at each STA as well as the amount of data load, thus combining the benefits of both the RSSI-based scheme [34] and load-based selection scheme [36]. On the contrary, the RSSI based scheme ignores the data load of the devices.…”