Gain-aware Content Dissemination in Social Wireless Networks


Recent emergence of Social Wireless Networks (SWNs) and Device-to-Device (D2D) communication approaches have opened up new ways of disseminating commercial content such as coupons and electronic viral marketing. SWNs, which are special type of Delay Tolerant Networks (DTNs), can be formed in public places such as in a university campus or in a work place, where individuals carrying mobile devices form D2D ad hoc networks. These networks often reflect inter-personal social ties, interactions, and similarities in content and commodity consumption patterns, which can be leveraged for D2D commercial content (e.g., coupon) dissemination with the goal of maximizing a predefined commercial gain for a Content Provider (CP). In other words, a carrier of content/coupon identifies most fit individual Content Consumers (CCs), based on its pattern of social interaction (e.g., contact duration, centrality, etc.) with target consumers.
DTN Content Dissemination in Social Wireless Networks

In D2D distribution, in comparison to the traditional centralized content dissemination methods, the primary cost comes from incentivizing mobile nodes for taking part in the dissemination process. D2D coupon forwarding costs bandwidth, storage, CPU and, more importantly, battery power which can deter people from taking part in dissemination. Considering above mentioned parameters, the final goal is to come up with a practical solution which enables D2D coupon dissemination possible in real world scenarios. As a step towards that, the project involved developing an Android based phone application for collecting users' interaction and product interest profiles via Bluetooth links. That is the base platform upon which one of the designed content dissemination algorithms will be implemented for real world test scenarios within SWNs such as Michigan State University campus.


Objective


The objective of this project is to develop coupon dissemination mechanisms that can maximize a coupon’s economic gain in the presence of costs/rebates for D2D forwarding incentives. This is done given a combination of commercial parameters (e.g., earned revenue , discount , rebate , distribution upper bound , expiry time , redemption function) and network parameters (e.g., social wireless network of consumers with various interest profiles and interaction profiles).


Proposed Mechanisms


The developed architectures in this work propose two different definitions of economic gain, one as a function of coupon delivery delay [6-7] and another, a gain utility [3] that is designed around the notion of consumption interest and coupon redemption probability. Different solutions have been investigated with the goal of maximizing such economic gain for the coupon generator. These solutions include heuristic, history based, predictive methods [1,3,6,7], as well as reinforcement learning (Q-learning) based approaches [2]. Ongoing research on this topic includes studying the feasibility of leveraging evolutionary protocol synthesis [4-5] for content dissemination in SWNs and embedding LSTM/Neural Network based prediction mechanisms for prediction of the evolving economic gain.


Selected Publications


[1] F. Hajiaghajani and S. Biswas, Device-to-Device Commercial Content Dissemination in Social Wireless Networks, 14th IEEE Annual Consumer Communications & Networking Conference (CCNC), Las Vegas, US, Jan 2017.

[2] F. Hajiaghajani and S. Biswas, Learning based Gain-aware Content Dissemination in Delay Tolerant Networks, 9th International Conference on COMmunication Systems & NETworkS (COMSNETS), Bengaluru, India, Jan 2017.

[3] F. Hajiaghajani and S. Biswas, Device-to-Device Coupon Distribution using Economic Routing Utilities , 12th IEEE International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob) 2016, New York, US, Oct. 2016.

[4] F. Hajiaghajani and S. Biswas, Feasibility of Evolutionary Design for Multi-access MAC Protocols, In proceedings of IEEE Globecom 2015, San Diego, December 2015.

[5] F. Hajiaghajani and S. Biswas, MAC protocol Synthesis using Evolvable State-Machines, In Proceedings of 24th International Conference on Computer Communications and Networks (ICCCN), Workshop on wireless mesh and Ad-hoc networking, Las Vegas, US, August 2015.

[6] F. Hajiaghajani, Y. Piolet, M. Taghizadeh, and S. Biswas, Economy Driven Content Dissemination in Delay Tolerant Networks, Elsevier Ad Hoc Networks, Volume 20, September 2014, Pages 132–149.

[7] F. Hajiaghajani and S. Biswas, Gain-aware Multicast Content Routing in Delay Tolerant Networks, SPIE Defense, Security, and Sensing Symposium, 29 April - 3 May 2013, Baltimore, Maryland.