روش های جدید برای تجزیه و تحلیل عملکرد ARQ تکراری انتخابی هیبریدی تطبیقی ​​

In this paper we present two novel approximation methods for the performance analysis of selective repeat (SR) ARQ protocol with any size of buffer. The methods are called sequential method (SM) and stable function method (SFM). Both SM and SFM methods provide highly accurate results as compared to other conventional methods of analysis. However, between SM and SFM methods, it is observed that SM is computationally more complex. Hence, SFM method is proposed for our analysis. Based on SFM, we present an ad hoc algorithm for an adaptive SR ARQ scheme to optimize the performance of SR ARQ. The performance of hybrid SR ARQ (an SR ARQ protocol combined with FEC) is also analyzed. It is concluded from a comparison of different methods of analyzing the SR ARQ protocol that the stable function method (SFM) provides a simple and novel approach for network analysis.

Real-World data communication is subject to errors caused by various sources. Protocol for achieving efficient and reliable communication between source and receiver is very important to ensure the communication quality. One widely used technique for handling errors at the data-link layer of data communication systems is the error detection incorporated with automatic-repeated-request (ARQ). The protocol may be categorized as: 1. Stop-and-wait (SW); 2. Go-back-N; 3. Selective-repeat (SR). Much work has been done for analyzing these basic protocols [1], [3] and [6]. Generally speaking, the SW protocol suffers from certain inefficiency due to the channel being idle between the transmission of the message and the reception of the ACK/NACK from the receiver. This inefficiency is particularly serious when the round-trip delay between the transmitter and the receiver is long compared to transmission time of a message. SR ARQ offers the best performance in terms of throughput among the three, although it has an increased requirement for buffers at the receiver. The buffers are needed to store those packets which are received out of sequence. Variations of the basic SR ARQ protocol have also been extensively studied [4], [5] and [17]. In systems where the packet lengths are relatively large, and where the noise and/or interference levels are high, SR ARQ with only error detection results in a low throughput due to the large number of retransmissions required. Satellite networks and packet radio systems are examples of such systems [7] and [8]. In these instances, a combination of error detection and error correction can offer significant advantages over an error detection only system. This is called hybrid ARQ schemes. Hybrid ARQ schemes have been widely studied [2], [16] and [18]. Generally, the analysis on throughput efficiency of SR ARQ protocols focuses on the case of infinite buffer [2] and [18], or on lower bound methods by approximation [1] and [9]. In the case of analysis with finite buffer, the buffer size is mostly assumed to be the number of times of W packets [1] and [9], where W is the number of packets that can be sent in a round-trip time. The performance analysis of SR ARQ protocol with different buffers appeared for the first time in Ref. [1]. The analysis provides only a lower bound performance. In the case of W being small, the above analysis methods can often provide a reasonable throughput efficiency. However, in the case of broadband satellite networks and packet radio system, W is very high [7] and [8]. For a simple comparison, the throughput efficiency of Go-back-N ARQ is 1/(1+W∗(ρ/1−ρ)), while the throughput efficiency of SR ARQ with infinite buffer is 1−ρ, where ρ is packet error ratio. When W is large, the difference on throughput efficiency between Go-back-N and SR ARQ with infinite buffer is also very large. In such cases, efficient analytical methods are necessary for SR ARQ protocols to investigate the impact of the receiver buffer. In this paper we estimate the throughput efficiency of SR ARQ protocol with any finite buffer. The buffer size can vary from 1 to ∞. When buffer size is set to 1, the SR ARQ protocol becomes Go-back-N ARQ. The paper is organized as follows. In 2 and 3, we propose a sequential method (SM) and a stable function method (SFM) for accurate analysis on throughput efficiency. In Section 4, we present an algorithm to optimize the performance of pure SR ARQ protocol by continuously sending an optimum number of copies of a packet. The optimum number is adaptively changed with the physical conditions. In Section 5, we analyze the performance of a hybrid SR ARQ, which combines a simple concatenated coding system. SFM is compared with two conventional methods in Section 6. Numerical results on the comparison of the performance of pure SR ARQ with those of optimum SR ARQ and hybrid SR ARQ are also presented in this section. Section 7 provides our conclusions.

This paper presents two novel methods, sequential method (SM) and state function method (SFM), for calculating the throughput of pure SR ARQ protocol with any finite buffer. The methods are especially useful for a large bandwidth-delay product case, such as broadband satellite communications. When the transferred file is large, SM and SFM have similar accuracy in calculating the throughput. The two methods show accurately how the throughput is related to packet error ratio, and receiver buffer size B. The receiver buffer size B can be 1 to infinite. SFM not only simplifies the calculation of the throughput, but also yields a higher accuracy compared to other two conventional analytical methods. An optimum scheme of SR ARQ is often applied for improving the performance of SR ARQ by sending an optimum number of copies. Based on SFM, we present an ad hoc algorithm to determine the optimum value for the optimum SR ARQ protocol. SFM simplifies the analysis on the optimum scheme. Based on SFM, we also analyze a kind of hybrid SR ARQ scheme, in which a concatenated coding system is applied for error correction. Numerical results show that the improvement obtained from the HSR ARQ scheme on the performance is more evident than that from the optimum SR ARQ scheme. Extending SFM to analyze the performance of multi-receiver SR ARQ protocol will be our future work