MC-UWMAC: A multi-channel MAC protocol for underwater sensor networks

Chaima Zidi, Fatma Bouabdallah, Raouf Boutaba, Ahmed Mehaoua

TL;DR

This paper introduces MC-UWMAC, a novel multi-channel MAC protocol designed for underwater acoustic sensor networks (UW-ASNs). It addresses the unique challenges of underwater communication by employing a single control channel for handshaking and multiple data channels for efficient data transfer, significantly improving network performance in terms of energy efficiency, packet delivery ratio, and end-to-end delay.

What Problem Does This Paper Address?

Underwater acoustic sensor networks (UW-ASNs) face significant challenges due to the inherent limitations of acoustic propagation, including low speed, high latency, and significant energy consumption. Traditional terrestrial MAC protocols are ill-suited for this environment. Existing UW-ASN MAC protocols often struggle with:

  • Energy Efficiency: Underwater sensors are typically battery-powered and difficult to recharge, making energy conservation paramount. Collisions and unnecessary retransmissions are major energy drains.
  • Collision Avoidance: The underwater acoustic channel is prone to interference and signal degradation, increasing the likelihood of collisions. This is exacerbated in multi-hop networks where hidden terminal problems are common.
  • Throughput and Latency: The low speed of acoustic communication and the need for robust error correction mechanisms lead to high latency and limited throughput.
  • Multi-channel Communication: While multi-channel communication offers the potential for increased throughput, effectively managing channel access and avoiding collisions across multiple channels, especially with single-transceiver nodes, is complex.

This paper aims to bridge this gap by proposing a MAC protocol that specifically tackles these issues in the context of UW-ASNs.

Methods

MC-UWMAC is a multi-channel MAC protocol that operates on a single control channel and multiple data channels. Its key innovations lie in:

  • Single Control Channel with Virtual Grid Slot Assignment: A common slotted control channel is used for handshaking (RTS/CTS). To ensure collision-free access, nodes are assigned unique slots within this control channel based on a virtual grid partitioning of the network area. This assignment is calculated using the node's geographical coordinates and a parameter 'p' related to the grid size, ensuring that nodes in adjacent "small squares" are unlikely to use the same slot. This avoids the need for explicit neighbor discovery or extra message exchanges for slot allocation.
  • Quorum-Based Data Channel Allocation: For data transmission, MC-UWMAC utilizes multiple data channels. A unique data channel is allocated to each pair of communicating neighbors. This allocation is based on a "singleton-intersecting quorum system," which guarantees that each pair of neighbors has a distinct data channel for their communication. This mechanism effectively avoids the multi-channel hidden terminal problem and ensures that data channels are used efficiently and without collisions.
  • Simultaneous Control and Data Communication: The protocol allows for simultaneous communication on the control channel (for handshaking) and data channels. This means that once a handshake is successful, nodes can immediately switch to their assigned data channel, reducing idle time and improving overall network efficiency.

Datasets

Not discussed in the paper. The evaluation is based on simulations.

Key Insights and Ideas

  • The "Virtual Grid" for Slot Assignment: The clever use of a virtual grid and coordinate-based slot assignment is a key insight. It allows for decentralized, collision-aware slot allocation without requiring explicit neighbor discovery, which is crucial in dynamic underwater environments.
  • Singleton-Intersecting Quorum Systems for Channel Allocation: Applying quorum systems, specifically singleton-intersecting ones, to data channel allocation is a novel approach. It provides a mathematically sound way to guarantee unique channel assignments between neighbors, effectively eliminating multi-channel hidden terminal issues.
  • The Trade-off with Parameter 'p': The paper highlights a critical trade-off related to the parameter 'p' (which influences the grid size and number of slots). Increasing 'p' reduces collisions on the control channel but can increase end-to-end delay and energy consumption due to smaller data channel bandwidth. Finding the optimal 'p' is essential for balancing these factors.
  • MC-UWMAC's Superiority in UW-ASNs: The results strongly suggest that MC-UWMAC's design principles are well-suited for the unique constraints of underwater acoustic networks, outperforming existing protocols like MM-MAC.

Key Results

The paper presents simulation-based results demonstrating the effectiveness of MC-UWMAC:

  • Packet Delivery Ratio: MC-UWMAC consistently achieves a higher packet delivery ratio (above 90%) compared to MM-MAC, with improvements of up to 20%. This is attributed to its robust collision avoidance mechanisms.
  • End-to-End Delay: MC-UWMAC exhibits lower end-to-end delay, especially at higher traffic rates. This is partly due to the ability to switch to data channels immediately after handshaking, without waiting for a dedicated data period.
  • Energy Consumption: MC-UWMAC is shown to be more energy-efficient, consuming less energy per useful bit. This is a direct consequence of reduced collisions and more efficient channel utilization.
  • Collision Probability: The collision probability on the control channel is inversely proportional to the parameter 'p'. Increasing 'p' leads to fewer collisions but can impact other performance metrics. The paper suggests 'p=3' as an optimal value for the studied network model.

Limitations and Future Work

  • Collision Probability on Control Channel: While MC-UWMAC significantly reduces collisions, unlikely collisions can still occur on the control channel if multiple nodes reside in the same "small cell" and attempt to transmit RTS packets simultaneously. The paper acknowledges this and proposes a back-off mechanism to handle such situations.
  • Parameter 'p' Optimization: The optimal value of 'p' is network-dependent and requires careful tuning. The paper suggests that 'p' depends on the network topology and maximum neighborhood size.
  • Future Work: The authors suggest that further research could explore more sophisticated back-off mechanisms for control channel collisions and investigate the protocol's performance in more complex underwater scenarios with varying node densities and mobility patterns. The paper also implies that further optimization of the quorum system and slot assignment could lead to even greater efficiency.

Summary updated

Abstract

Fundamental differences between underwater acoustic propagation and terrestrial radio propagation impose the design of new networking protocols. In this paper, a multichannel MAC protocol, MC-UWMAC, especially designed for underwater acoustic sensor networks, is proposed and evaluated. MC-UWMAC is a low power MAC protocol operating on multichannel using a single slotted control channel and multiple data channels. To guarantee a collision free communication, MC-UWMAC uses a virtual grid based slot assignment linked with a quorum based data channel allocation. Specifically, control channel slots are dedicated for handshaking. Data transmission takes place in a unique data channel especially reserved for each communicating pair. Simulation results show that MC-UWMAC can greatly improve the network performance especially in terms of energy consumption, packet delivery ratio and end-to-end delay.
BibTeX
@inproceedings{zidi2017mc,
  author = {Chaima Zidi and Fatma Bouabdallah and Raouf Boutaba and Ahmed Mehaoua},
  title = {{MC-UWMAC: A multi-channel MAC protocol for underwater sensor networks}},
  booktitle = {WINCOM},
  year = {2017},
  doi = {10.1109/wincom.2017.8238152},
}
Figure