Energy-Aware Social-Based Multicast in Delay-Tolerant Networks

EASDM: Balancing Social Influence and Battery Life in Delay-Tolerant Networks

Energy-Aware Social-Based Multicast in Delay-Tolerant Networks

2015-05-01

Animesh Roy, Tamaghna Acharya, Sipra Das Bit

Summary

Problem

Method

Results

Takeaways

Abstract

This paper introduces EASDM (Energy-Aware Single-Data Multicast), a relay selection scheme for Delay-Tolerant Networks (DTNs) that integrates social centrality with residual battery life. It addresses the limitation of previous social-based routing by balancing delivery success with network longevity, outperforming the standard SDM baseline in sustained multicast operations.

    ## TL;DR
    In the unpredictable world of Delay-Tolerant Networks (DTNs), "popular" nodes—those with high social centrality—are often worked to death. This paper proposes **EASDM**, a routing scheme that selectively "rests" influential nodes by considering their residual energy. By treating relay selection as a constrained optimization problem, the authors achieve a **27% increase in supported multicast sessions** compared to state-of-the-art social-only methods.

    ## The Popularity Trap: Why Social Routing Fails
    Current DTN routing strategies, such as the Single-Data Multicast (SDM) scheme, typically rely on **Centrality**. Centrality measures a node's likelihood of encountering others (Cumulative Contact Probability - CCP). 
    
    The logic is sound: if a node meets many people, it is a great courier. However, this creates a **systemic bottleneck**. These high-centrality nodes are selected repeatedly across every multicast session. In energy-starved scenarios like disaster relief, these "social butterflies" run out of battery first, leaving the network fragmented and "dead" despite having many low-centrality nodes still active.

    ## Methodology: Greed is Good (When It’s Energetic)
    The authors redefine relay selection not just as a search for the best path, but as a **Multi-Objective Optimization** problem.
    
    ### 1. The Optimization Goal
    The goal is to minimize the number of relays ($k$) while maximizing the sum of their residual energy ($E_i$), subject to:
    *   Achieving a target delivery ratio ($p$).
    *   Ensuring every relay has energy above a minimum threshold ($e_0$).

    ### 2. The Heuristic: EASDM Algorithm
    Since solving this NP-hard problem optimally in a resource-constrained DTN is impractical, the authors use a greedy approach:
    *   **Partitioning**: Nodes are split into High-Centrality and Low-Centrality sets.
    *   **Reservation Strategy**: A portion ($\alpha$) of the high-centrality nodes is intentionally **reserved** for future sessions, even if they are available now.
    *   **Energy-First Selection**: Within the active sets, nodes are sorted by residual energy rather than just their contact probability.

    ![EASDM Relay Selection Logic](https://cdn.atominnolab.com/wisdoc/images/20260603-4753b790-04f6-4430-865e-7b145fb672a9/page_003_block_000.png)
    *Figure: Comparative relay selection between SDM (Pure Social) and EASDM (Energy-Aware).*

    ## Performance and Results
    The researchers simulated a Bluetooth-based DTN (41 devices) with specific energy hardware profiles (432mW for TX, 425mW for RX).

    *   **Network Longevity**: EASDM consistently supported more multicast sessions. At high delivery requirements, EASDM maintained a 27% lead over SDM.
    *   **The Cost of Sustainability**: To achieve the same delivery ratio without exhausting high-centrality nodes, EASDM occasionally uses "sub-optimal" relays. This leads to a slight increase in total relay count (up to 11%), which is a small price to pay for a 14% higher survival rate of critical nodes.

    ![Experimental Results](https://cdn.atominnolab.com/wisdoc/images/20260603-4753b790-04f6-4430-865e-7b145fb672a9/page_004_block_004.png)
    *Figure: Comparison of supported multicast sessions vs. required delivery ratio.*

    ## Critical Insights
    The core "magic" of EASDM isn't just checking battery levels; it's the **Reservation Parameter ($\alpha$)**. By acknowledging that a high-centrality node is a precious resource that shouldn't be "spent" all at once, the algorithm introduces a form of **load-balancing** across time rather than just space.

    ### Limitations and Future Work
    *   **Dynamic Alpha**: Currently, $\alpha$ is a fixed fraction. In a real-world deployment, this should likely be dynamic, scaling with the urgency of the data.
    *   **Mobile Fairness**: The model assumes all nodes start with equal energy. Future studies should investigate how EASDM handles "rogue" nodes or nodes that start with significantly lower power.

    ## Conclusion
    EASDM proves that in intermittent networks, **sustainability is a social contract**. By sharing the burden of data forwarding among energy-rich nodes—even those with lower social standing—the network as a whole remains functional for significantly longer periods.

Find Similar Papers

Try Our Examples

Search for recent papers that utilize Reinforcement Learning to dynamically adjust the reservation parameter alpha in energy-aware DTN routing.
Which paper first proposed the Cumulative Contact Probability (CCP) metric for social-aware multicast, and how does EASDM's energy constraint mathematically alter its objective function?
Examine how the EASDM relay selection strategy can be adapted for heterogeneous energy models in Underwater Wireless Sensor Networks (UWSNs).

Contents

EASDM: Balancing Social Influence and Battery Life in Delay-Tolerant Networks

1. TL;DR

2. The Popularity Trap: Why Social Routing Fails

3. Methodology: Greed is Good (When It’s Energetic)

3.1. 1. The Optimization Goal

3.2. 2. The Heuristic: EASDM Algorithm

4. Performance and Results

5. Critical Insights

5.1. Limitations and Future Work

6. Conclusion