Abstract: Wireless LANs are currently being studied as vehicle-to-ground communication systems for urban rail transit onboard monitoring systems and passenger information systems. To ensure that the throughput, bit error rate, transmission delay, and anti-interference capabilities provided by wireless LANs meet the requirements of vehicle-to-ground communication, this paper simulates different rail transit communication environments with two mobile stations and multiple mobile stations, and analyzes and discusses the performance of wireless LANs. By optimizing the MAC layer parameters, the medium access delay and collisions during transmission are effectively reduced, thereby improving the network throughput. Keywords: Wireless LAN; Media access delay; 802.11 [b][align=center]WLAN Performance Analysis and Optimization in Metro Line Communication Cai Wei-hao, Zheng Guo-xin, He Hui, Zhou Xiang-wei[/align][/b] Abstract: At present, WLAN is being researched as the train-ground communication system for metro line onboard surveillance system and passenger information system. To meet the demands of train-ground communication for throughput, BER, transmission delay, and anti-jamming ability provided by WLAN, we simulate different metro line communication environments of two mobile stations and multiple mobile stations, analyze and discuss WLAN performance. By optimizing the MAC layer parameters, we effectively reduce the media access delay and collision, thus improving the network throughput. Key words: WLAN; media access delay; 802.11 1. Introduction Rail transit has become an ideal means of transportation in modern cities. Various control and service systems, such as communication-based train control systems (CBTC), are used in various applications. Control), train video surveillance system, passenger information system (PIS), multimedia services, etc. all use wireless local area network (WLAN) to complete data transmission. IEEE 802.11 is the standard of WLAN. It is currently mainly used for small-scale fixed wireless network applications, such as wireless Internet access. When it is applied to the ground communication of rail transit, due to the transmission of data by electromagnetic waves in space, its anti-interference ability is poor, but it has problems such as large medium access delay, high bit error rate, and low throughput. In order to improve the throughput of wireless local area network and reduce medium access delay, Reference [1] proposes to use different contention window parameters for different access applications to improve network throughput and ensure QoS. Reference [2] analyzes the competition fairness problem in wireless local area network. This paper optimizes the contention window, SIFS (Short Inter-Frame Space) and slot time, and uses OPNET to perform software simulation of the rail communication environment under different contention intensities. Through analysis, it proposes an optimization method combining contention window and SIFS to improve the performance of wireless network and enable wireless local area network to be better applied in rail transit. The second part of this paper introduces the application of wireless local area networks (WLANs) in urban rail transit. The third part simulates and analyzes the results of track communication between two mobile stations. The fourth part simulates random communication between multiple mobile stations in a track communication environment. The fifth part presents contention windows and SIFS optimization methods for different contention intensities. 2. Application of Wireless Local Area Networks in Urban Rail Transit Currently, WLANs have been applied to the CBTC train control system on Shanghai Metro Lines 6, 9, and 10. The CBTC system can perform real-time train movement permission updates via wireless network, shortening train intervals, increasing train density, and improving transportation efficiency. WLANs are also applied to the PIS passenger information system and real-time video monitoring system, providing passengers with necessary information and enabling video monitoring of train carriages, avoiding blind spots in track monitoring. The CBTC train control system and the wireless video monitoring system adopt similar system structures. The difference is that the CBTC system has onboard axle recorders, transponders, and other train signal control equipment to manage train control information. The video monitoring system, on the other hand, has onboard cameras, video encoders, and onboard server data processing equipment to process train video information. In the CBTC system, the onboard server transmits train information via onboard antennas to access points on both sides of the track, and then to the control center via a wired network. The control center then uses trackside signals, track circuits, and switch machines to reliably, efficiently, safely, and accurately control the train. In the video surveillance system, the onboard video acquisition system encodes and channels the images in real time. The onboard server transmits the data from multiple onboard antennas to trackside wireless access points via a wireless LAN. After processing by the trackside server, the data is transmitted to the control center via a ground wired network, enabling video monitoring of the train. The main difference between the two systems is the onboard equipment and the different processing at the control center. The data transmission via the wireless LAN and the data forwarding via the wired network are essentially the same. 3. Channel Analysis and Parameter Optimization under Dual Mobile Stations Due to the characteristics of subway train-to-ground wireless communication, the competition generated by the multiple transceiver antennas in the onboard network increases the latency and decreases the throughput of the wireless network medium access, failing to meet the bandwidth and latency requirements of the video surveillance system. First, the transmission process of the network is analyzed using dual mobile stations. The IEEE 802.11 MAC layer provides two access mechanisms: Distributed Coordinated Operation (DCF) and Point Coordinated Operation (PCF). DCF is based on Carrier Sense Multiple Access (CSMA/CA) technology with collision avoidance. Before transmitting data, the mobile station listens to see if other stations on the same channel are transmitting data. If other stations are transmitting data, it continues to listen. Once the channel is idle, it enters a backoff process. After the backoff process ends, it transmits data. Because a mobile station cannot listen to whether its channel is in conflict while transmitting data in 802.11, a collision avoidance (CA) mechanism is used instead of a collision detection (CD) mechanism. When the destination mobile station successfully receives the data, it sends an ACK frame. The source mobile station acknowledges successful transmission upon receiving the ACK frame; otherwise, it considers a collision to have occurred and retransmits the data, thus avoiding a collision. This process is called Automatic Repeat Request (ARQ), and it is often used when errors such as receiver noise, electromagnetic interference, spatial movement, and data collisions cause abnormal ACKs. IEEE 802.11 supports three different types of frames: management frames, control frames, and data frames. Management frames are used for connection and disconnection between stations and access points, timing and synchronization, and authentication. Control frames are used for handshake communication and positive acknowledgment during contention periods and for ending non-contention periods. Data frames are used to transmit data during both contention and non-contention periods. In a contention-driven environment, different types of frames have different priorities. 802.11 distinguishes the priorities of different frames by using different inter-frame spacings (IFS). The inter-frame spacing is the interval between frames. 802.11 defines several different IFS: SIFS, PIFS, DIFS, and EIFS. Slot Time is the time required to transmit the maximum theoretical distance between two nodes. PIFS = SIFS + Slot Time; DIFS = SIFS + 2 Slot Time; EIFS = SIFS + 8 ACK + Preamble Length + PLCP Header Length + DIFS. Control frames such as RTS, CTS, and ACK have a SIFS inter-frame spacing, which is the shortest and therefore has the highest priority. PIFS is used for data frames with high real-time requirements and has the next highest priority. DIFS is used for general asynchronous data transmission frames and has the next lowest priority. EIFS is used only when MAC protocol data unit transmission fails and is retransmitted, and has the lowest priority. In 802.11, channel contention causes data collisions, leading to data transmission failure. DCF mode shares the channel through contention, using control frames such as RTS, CTS, and ACK. As shown in Figure 1, in CSMA/CA in DCF, the mobile station first listens to the channel before sending data. If the channel is busy, it continues to listen. If the channel remains idle for one DIFS period, the mobile station enters a dynamic backoff process. Before each data transmission, the mobile station randomly selects an integer Backoff Time as the initial value of the backoff counter for this transmission process. Backoff Time is the backoff time that the mobile station must wait before starting to transmit data. Backoff Time = Random * Slot Time, where Random is a pseudo-random number distributed in [0, CW], and CW is taken from the minimum contention window CWmin and the maximum contention window CWmax. In the first data transmission, CW is CWmin. If the data transmission fails and is retransmitted, CW increases exponentially by 2 until CWmax. If the channel remains idle, the mobile station decrements its backoff counter by 1 every Slot Time. A backoff counter of 0 indicates the backoff process is complete, and the station can immediately begin sending RTS control frames or data frames. If the channel is occupied during the backoff process, the mobile station stops decrementing the backoff time and waits for the channel to remain idle for DIFS and the next Slot Time before resuming decrementing. [align=center]Figure 1 DCF Backoff Mechanism[/align] The PCF method uses a polling mechanism, where each node can only send and receive data after being polled, thus avoiding contention. DCF is more efficient and is the primary method in 802.11, while PCF is less efficient and more complex to implement, usually only an optional method. The simulation tool used is OPNET. The mobile station model uses frequency hopping spread spectrum physical layer modulation, and its MAC layer specifies fixed values ​​for parameters such as Slot Time, SIFS, CWmax, and CWmin. By customizing the MAC layer processing model and changing these parameters to variables, the network model was optimized. First, a dual-mobile-station track communication environment is constructed. The data source parameters of the mobile stations are set according to Table 1, generating a data stream of approximately 850kbps, with data being sent and received between them. The simulation time is 80 seconds. Table 1: Data Source Parameters. During the simulation, the average media access latency of the mobile stations is collected for network performance analysis. Media access latency refers to the total time from when a data packet enters the queue to when it is sent by the physical layer. The shorter the latency, the better the network performance. Three simulation schemes are used: a. Simulating the average media access latency by reducing Slot Time and SIFS, and analyzing the impact of Slot Time and SIFS on network performance. b. Reducing the minimum contention window CWmin. Reducing CWmin also reduces the backoff time of the mobile stations. c. Simultaneously reducing CWmin, Slot Time, and SIFS, and analyzing network performance. Specific parameter settings are shown in Table 2. Table 2 MAC Parameter Settings [align=center] Figure 2 Average Media Access Latency of Two Mobile Stations Figure 3 Average Throughput of Two Mobile Stations[/align] The average media access latency was simulated, and the results are shown in Figure 2. The horizontal axis represents the simulation time, and the vertical axis represents the latency, both in seconds. When the network load is light and contention is low, the average media access latency of the standard 802.11 is 0.38ms. Reducing the Slot Time and SIFS reduces the average media access latency to approximately 0.23ms. If CWmin is reduced, i.e., the contention window is reduced, the latency is approximately 0.3ms. When CWmin, Slot Time, and SIFS are reduced in combination, the average media access latency drops to approximately 0.19ms, half of the original, improving network performance. The average throughput of the mobile stations was simulated using the above methods, as shown in Figure 3. Throughput refers to the total amount of data received by the upper layer from the MAC layer. The simulation results show that the average throughput of the four methods is basically the same, approximately 800kbps. When the network is under low load, using a combination of methods to reduce CWmin, Slot Time, and SIFS can reduce the average media access delay, but cannot improve the average network throughput. 4. Channel Analysis and Parameter Optimization under Multiple Mobile Stations The multi-mobile-station rail transit simulation environment consists of 8 mobile stations. Each mobile station randomly selects a destination mobile station to send data at a data rate of 850kbps. The simulation simulates the network state under high network load. The performance metrics of the mobile stations are collected for network performance analysis. [align=center] Figure 4 Average Media Access Delay of Multiple Mobile Stations Figure 5 Average Throughput of Multiple Mobile Stations[/align] With the parameter settings the same as in the dual-mobile-station environment, the simulation results of the average media access delay are shown in Figure 4. Due to the increased number of mobile stations, the contention in the channel caused by the random transmission of data packets by each mobile station also increases, resulting in a significant increase in average media access delay compared to the dual-mobile-station environment. The standard average media access latency is approximately 0.42s. Reducing Slot Time and SIFS latency reduces it to 0.25s, while reducing CWmin latency is approximately 0.38s. The improvement in media access latency is not significant, and is less effective than reducing CWmin in a dual-mobile-station environment. Simultaneously reducing CWmin, Slot Time, and SIFS latency is approximately 0.29s, which actually increases latency compared to reducing Slot Time and SIFS. This is because in a high-contention environment, reducing the minimum contention window (CWmin) shortens the backoff time, leading to repeated packet collisions and degrading network performance. Simulation results for average throughput are shown in Figure 5. The simulation results show that in a multi-node contention environment, the standard average throughput is approximately 580kbps. Reducing Slot Time and SIFS increases the average throughput to 640kbps, improving the network's average throughput. However, reducing CWmin results in an average throughput of 550kbps, a decrease compared to the standard throughput. Reducing the minimum contention window (CWmin) decreases the backoff time for each mobile station when the channel is busy. However, due to increased contention among multiple nodes, the reduced backoff time for each mobile station leads to a decrease in average throughput. The two simulation results show that reducing CWmin is ineffective in improving network performance under high contention conditions. 5. Conclusion This paper constructs two different rail transit wireless communication network models—one with two mobile stations and the other with multiple mobile stations—to analyze network performance under different contention intensities. In a low-contention rail environment, reducing CWmin, Slot Time, and SIFS significantly improves network latency and optimizes network performance. Under high contention, reducing CWmin exacerbates packet collisions and reduces throughput. Reducing Slot Time and SIFS reduces latency and increases throughput. Choosing different parameters for different network conditions and rail transit application characteristics can effectively improve the average medium access latency and average throughput of wireless LANs, thereby enhancing network performance. Wireless networks have broad development and application prospects in real-time monitoring, control, and multimedia information services in rail transit. The innovation of this paper lies in using software to analyze and optimize the performance of wireless LANs in rail transit. References 1 Aad I, Castelluccia C. Differentiation mechanisms for IEEE 802.11[J]. In: Proc. of the 20th Annual Joint Conf of the IEEE Computer and Communications Societies, 2001. 209-218 2 Ozugur T, Naghshineh M, Kermani P, Copeland JA. Fair media access for wireless LANs[J]. In: Proc. of the IEEE Global Telecommunications Conf. GLOBECOM. Rio de Janeiro: IEEE, 1999. 570-579. 3 Li Juan, Li Jun, Chen Yi, Yang Hongsheng. Research on MAC protocol of wireless LAN and implementation of real-time service QoS[J]. Microcomputer Information, 2006, 2-3: 155-157 4 Shu Anjie, Li Kaicheng. Research on information security transmission of CBTC system[J]. Microcomputer Information, 2006, 3-1: 44-46 5 Wang Wenbo, Zhang Jinwen. OPNET Modeler and Network Simulation [M]. Beijing: Posts & Telecom Press, 2003, pp. 25-97.