Introduction
This article provides a formal correction to the paper "Analysis of a High-Performance MAC Protocol for Underwater Acoustic Networks," originally published in the IEEE Journal of Oceanic Engineering (Vol. 39, No. 1, Jan. 2014). The original work established a framework for analyzing Medium Access Control (MAC) protocols in the challenging environment of underwater acoustic networks (UANs). This erratum specifically addresses a descriptive error in the timing logic of the MACA-based protocol model detailed in Section II-C.
The Challenge of Underwater Communication
Designing efficient MAC protocols for UANs is significantly more complex than for terrestrial radio networks. The primary hurdle is the extremely slow propagation speed of sound in water—approximately 1,500 m/s—which is five orders of magnitude slower than the speed of light. This high latency means that the "handshake" mechanisms used to avoid data collisions, such as the Request-to-Send (RTS) and Clear-to-Send (CTS) exchange, are highly sensitive to timing parameters.
In mathematical modeling and simulation of these protocols, precise definitions of when timers start and stop are critical. Even a small discrepancy in a timer's trigger point can lead to significant errors in calculating network throughput, energy efficiency, and collision probability. If a timer is defined incorrectly in a state machine model, the resulting performance analysis will not accurately reflect real-world behavior.
The Correction: Timer $t_A$
The correction pertains to Section II-C, "MACA-Based Protocol Model," specifically the third sentence of the second paragraph. The original text contained an ambiguity regarding the initiation of a key protocol timer, $t_A$, which governs the handshake process.
The revised sentence is as follows:
“Timer $t_A$ is started when RTS transmission begins.”
Significance of the Change
The distinction between starting a timer at the beginning of a transmission versus at the end (once the hardware has finished pushing the bits into the water) is vital in high-latency environments. In UANs, the duration required to transmit the RTS packet itself is non-negligible relative to the propagation delay.
By starting Timer $t_A$ at the onset of the RTS transmission, the protocol logic correctly accounts for the period during which the sender is occupied with transmission before it transitions to a waiting state for the CTS response. If the timer were started only after the transmission ended, the model would misalign the expected response window, potentially leading to premature timeouts or inefficient channel utilization. For researchers attempting to replicate the authors' results or implement this protocol in underwater modems, this distinction is the difference between a functional system and a continuous timeout loop.
Conclusion
While this correction involves a single sentence, it is essential for the technical integrity of the MACA-based protocol model. This clarification ensures that the theoretical framework aligns with the intended operational logic used in the original paper's performance analysis. Researchers and engineers are advised to update their models to reflect that Timer $t_A$ is triggered at the start of RTS transmission to ensure accurate simulation and implementation of high-performance underwater MAC protocols.
Summary updated