Bridging the Social and Wireless Networking Divide: Information Dissemination in Integrated Cellular and Opportunistic Networks

Jie Hu, Lie-Liang Yang, H. Vincent Poor, Lajos HanzoIEEE Access 2015 · DOI

Abstract

Mobile users increasingly share common interests in information such as traffic updates and news. However, traditional centralized cellular infrastructures are not optimized for disseminating this "Information of Common Interest" (IoCI) to large numbers of requesters. This paper explores the integration of cellular networks with opportunistic device-to-device (D2D) networks. By leveraging Social Network Analysis (SNA) and hybrid dissemination schemes, we demonstrate that mobile users can offload up to 58% of cellular traffic while achieving delivery ratios exceeding 90% and significantly reducing energy consumption.

Introduction: The Design Dilemma

The explosive growth of mobile data demand has created a bottleneck for traditional infrastructure-based systems. Specifically, disseminating the same data to many users—the IoCI problem—is inefficient via standard unicast or multicast. The authors identify three primary challenges:

  1. Intermittent Connectivity in rural areas with sparse base station (BS) coverage.
  2. Infrastructure Overload due to high data demand causing congestion.
  3. Inefficient Crowdsourcing in dense environments like stadiums where interference and scheduling delays degrade performance.
Growth of research publications in the cross-disciplinary area of social-aided wireless networking.

To address these, the study bridges the gap between social behavior and wireless networking, utilizing the inherent mobility and social interactions of users to facilitate data spread.

Methodology: Socially-Aware Modeling

The research employs a cross-disciplinary approach combining Social Network Analysis (SNA) with mathematical modeling.

Social Seeding and Centrality

By using SNA metrics such as Degree, Betweenness, and Closeness Centrality, the system identifies "influential" mobile users. These users serve as initial "seeds" who receive data from the base station and then disseminate it to their social contacts through opportunistic encounters.

Markov Chain Analysis

The performance of these networks is modeled using Markov chains tailored to the network scale:

  • Large-Scale Networks: In rural or emergency scenarios where mobility is the primary factor, a Continuous-Time Pure-Birth Markov Chain (CT-PBMC) models the "infection" rate of data across the population based on inter-contact durations.
  • Small-Scale Networks: In dense areas where wireless channel physics and resource scheduling are the bottlenecks, a Discrete-Time Pure-Birth Markov Chain (DT-PBMC) is utilized.

Hybrid Information Dissemination Scheme

For dense, small-scale environments, the authors propose a two-stage hybrid scheme:

  1. Base Station-Aided Single-Hop Multicast (BSSHM): The BS transmits data to a select group of seed users.
  2. Cooperative Multicast-Aided Spontaneous Dissemination: Seed users then utilize D2D communication (e.g., Bluetooth, WiFi Direct) to spread data to peers using Time-Division-Multiple-Access (TDMA) and Round-Robin scheduling.

Experimental Validation and Results

The models were validated using the InfoCom 2006 Mobility Trace (78 users) and synthetic mobility models (Random Direction, Random Way-Point).

  • Traffic Offloading: Real-world trace data showed that opportunistic D2D links could offload 58% of total tele-traffic from the cellular backbone.
  • Reliability: In large-scale scenarios, a delivery ratio of over 90% was achieved with just 20 mobile users before data expiration.
  • Efficiency: The hybrid scheme reduced energy consumption by approximately 70% and dissemination delay by 40.6% compared to traditional BS-only multicast.

Conclusion and Future Directions

Integrating social insights into wireless networking transforms random mobility into a structured asset for data dissemination. While the results are promising, future work must address user cooperation (incentivizing "selfish" nodes), privacy concerns in social learning, and more advanced SNA-aware resource scheduling.

Summary updated

BibTeX
@article{hu2015bridging,
  author = {Jie Hu and Lie-Liang Yang and H. Vincent Poor and Lajos Hanzo},
  title = {{Bridging the Social and Wireless Networking Divide: Information Dissemination in Integrated Cellular and Opportunistic Networks}},
  journal = {IEEE Access},
  year = {2015},
  doi = {10.1109/access.2015.2477307},
}
Figure