In the railway industry, high-definition surveillance systems not only bring more clear images and better results, but also bring greater system performance requirements, long-distance transmission bandwidth and storage issues. What kind of high-definition system can be suitable for the railway industry multi-level access, long-distance transmission, monitoring large-scale characteristics of the industry?
In the entire traditional railway industry video surveillance program, its center is the streaming media server, which is common in the architecture of the station monitoring system. The front-end encoder or the unicast stream output by the IPC is transmitted to the distribution/storage server over the network, and the distribution/storage server implements unicast to multicast, on-demand distribution, and on-demand storage. In the case of the full SD system, this solution can basically meet the needs of multi-user operations, but if this solution is applied to high-definition systems, the drawbacks of streaming media distribution methods will be highlighted:
Railway monitoring solution
1. The server scale increases in proportion to the number of monitoring points. In particular, the high-definition IPC is expected to have better application effects. The output stream is generally 6M-8M, which is 3-4 times that of the conventional SD, and the distribution/storage server The number will increase greatly and the server must be upgraded because the performance of the server becomes a performance bottleneck in the solution.
2. There are many local failure points in the system. When one of the distribution/storage servers fails, it will affect the storage and real-time monitoring of multi-channel video images it manages. It does not meet the high security standards of the transportation industry. Reliability requirements. Although it is possible to solve this problem through redundant servers, the configuration of redundant servers at each station is considered from the cost point of view and is rarely used in actual projects.
3. The front-end IPC outputs a unicast stream, which causes the storage stream and the real-time stream to occupy the same bandwidth. If the high quality of the HD system is to be achieved, the disk array will be under greater pressure, and the cost will increase significantly. If it is to reduce the pressure of disk costs, increase the compression ratio of the storage flow, it will greatly reduce the clarity of high-definition surveillance images.
4. A large number of servers and their equipment racks not only occupy a large amount of equipment space, do not meet the requirements of green environmental protection, and are even more deadly to increase the unreliability of the system and increase the number of points of failure. When faced with emergencies and multiple hotspots, massive bursts of traffic can easily lead to paralysis of the entire server farm.
It represents a more advanced video surveillance solution. It uses a mature softswitch architecture in the telecom field: the NGN architecture. The management server is only responsible for handling device management, signaling distribution, and session establishment, and does not participate in real-time traffic flow (real-time streaming and storage streaming). In Figure 2, both the storage stream and the real-time stream are directly transmitted to the disk array, decoder, or monitoring client through the network, thus solving the problem of previous service performance and reliability bottlenecks.
When there are multiple users accessing the same-channel real-time video at the same time, multicast support is provided through the network, as indicated by the video stream labeled 1 in FIG. 2: the decoder and the monitoring client simultaneously access the same road image. Because the front-end IPC outputs a multicast stream, the switches in the network provide similar code stream replication functions, and solve the problem of performance and bandwidth impact caused by large-scale simultaneous access to the front-end IPC. This solution brings more advantages in HD systems:
1. There is no strong correlation between the size of the monitoring points and the number of servers. The video management server only participates in the process of establishing the session and does not participate in the distribution of video streams. Therefore, there is no system bottleneck problem in the system. A single video management server can meet the standard-definition, high-definition image monitoring of several thousand channels.
2. Due to the adoption of advanced softswitch NGN (service flow and control flow separation) architecture, failure of the management server will not interrupt the monitoring of real-time streaming, and the storage flow will not be affected. This ensures that the monitoring data is still recorded from the architecture. On the great guarantee of the system's high reliability, the entire system has no local failure point.
3, front-end IPC output dual stream (storage stream and real-time stream), two-channel video stream can be individually set the size of the stream, can maximize the high-definition monitor brings the high-definition effect, while, can significantly reduce the storage The cost protects the user's investment.
4. When emergencies or accidents occur and multiple hot spots occur, large-scale and bursty traffic is copied and distributed by the switch through multicast technology, thus easily solving the problem of large-scale multi-users and satisfying the High-reliability scheduling requirements of enterprise managers in emergencies.
The scenario shown above is mainly applied to the monitoring within the station, and it inevitably faces the problem of remote shared access by the superior dispatch center. The traditional streaming media distribution solution should be able to easily achieve multi-level network monitoring. However, if only the softswitch NGN architecture is used, the multicast mode will encounter major problems in the railway industry wide area network and low bandwidth.
Due to the nature of streaming media forwarding, the traditional solution naturally has a multi-tier architecture. At this time, the streaming media server in the station is responsible for the distribution and storage, and is also responsible for the forwarding function. This will inevitably increase the server's performance pressure. However, this solution solves the problem that the railway wide area network cannot implement multicast and has limited bandwidth. It forwards the video images in the station to the upper-level streaming media server, and the upstream streaming server completes the distribution function, which solves the problem of multiple users accessing the same image at the same time.
Due to the multi-level monitoring of railways, the network often cannot enable the multicast function. If the front-end IPC unicast is sent directly to the upper-level platform, it will inevitably consume limited WAN bandwidth, and it is also limited by the IPC's distribution capability. Then, how to make the appropriate modification to the Fig. 2 method so that it can obtain the features of high reliability, large capacity, and high performance, and at the same time, it can adapt to the shared access of the railway industry wide area network?
In fact, the corresponding solution is also very simple, that is, drawing lessons from the experience of traditional programs, adding a media forwarding server to the monitoring platform at the station level can solve the above problem, as shown in Figure 4. However, it must be noted that the monitoring software platform at the station level must be capable of simultaneously supporting front-end video streaming unicast and multicast modes, and can implement automatic unicast/multicast switching functions according to different access requirements. Such a solution can adapt to the large-scale application of high-definition systems in the railway industry in the future, and can ensure the efficient and reliable operation of the entire monitoring system. It is believed that the high-definition monitoring system will surely be more widely used in the railway industry, and how to deal with the impact of high-definition monitoring systems on existing solutions is worthy of everyone's common attention.
In the entire traditional railway industry video surveillance program, its center is the streaming media server, which is common in the architecture of the station monitoring system. The front-end encoder or the unicast stream output by the IPC is transmitted to the distribution/storage server over the network, and the distribution/storage server implements unicast to multicast, on-demand distribution, and on-demand storage. In the case of the full SD system, this solution can basically meet the needs of multi-user operations, but if this solution is applied to high-definition systems, the drawbacks of streaming media distribution methods will be highlighted:
Railway monitoring solution
1. The server scale increases in proportion to the number of monitoring points. In particular, the high-definition IPC is expected to have better application effects. The output stream is generally 6M-8M, which is 3-4 times that of the conventional SD, and the distribution/storage server The number will increase greatly and the server must be upgraded because the performance of the server becomes a performance bottleneck in the solution.
2. There are many local failure points in the system. When one of the distribution/storage servers fails, it will affect the storage and real-time monitoring of multi-channel video images it manages. It does not meet the high security standards of the transportation industry. Reliability requirements. Although it is possible to solve this problem through redundant servers, the configuration of redundant servers at each station is considered from the cost point of view and is rarely used in actual projects.
3. The front-end IPC outputs a unicast stream, which causes the storage stream and the real-time stream to occupy the same bandwidth. If the high quality of the HD system is to be achieved, the disk array will be under greater pressure, and the cost will increase significantly. If it is to reduce the pressure of disk costs, increase the compression ratio of the storage flow, it will greatly reduce the clarity of high-definition surveillance images.
4. A large number of servers and their equipment racks not only occupy a large amount of equipment space, do not meet the requirements of green environmental protection, and are even more deadly to increase the unreliability of the system and increase the number of points of failure. When faced with emergencies and multiple hotspots, massive bursts of traffic can easily lead to paralysis of the entire server farm.
It represents a more advanced video surveillance solution. It uses a mature softswitch architecture in the telecom field: the NGN architecture. The management server is only responsible for handling device management, signaling distribution, and session establishment, and does not participate in real-time traffic flow (real-time streaming and storage streaming). In Figure 2, both the storage stream and the real-time stream are directly transmitted to the disk array, decoder, or monitoring client through the network, thus solving the problem of previous service performance and reliability bottlenecks.
When there are multiple users accessing the same-channel real-time video at the same time, multicast support is provided through the network, as indicated by the video stream labeled 1 in FIG. 2: the decoder and the monitoring client simultaneously access the same road image. Because the front-end IPC outputs a multicast stream, the switches in the network provide similar code stream replication functions, and solve the problem of performance and bandwidth impact caused by large-scale simultaneous access to the front-end IPC. This solution brings more advantages in HD systems:
1. There is no strong correlation between the size of the monitoring points and the number of servers. The video management server only participates in the process of establishing the session and does not participate in the distribution of video streams. Therefore, there is no system bottleneck problem in the system. A single video management server can meet the standard-definition, high-definition image monitoring of several thousand channels.
2. Due to the adoption of advanced softswitch NGN (service flow and control flow separation) architecture, failure of the management server will not interrupt the monitoring of real-time streaming, and the storage flow will not be affected. This ensures that the monitoring data is still recorded from the architecture. On the great guarantee of the system's high reliability, the entire system has no local failure point.
3, front-end IPC output dual stream (storage stream and real-time stream), two-channel video stream can be individually set the size of the stream, can maximize the high-definition monitor brings the high-definition effect, while, can significantly reduce the storage The cost protects the user's investment.
4. When emergencies or accidents occur and multiple hot spots occur, large-scale and bursty traffic is copied and distributed by the switch through multicast technology, thus easily solving the problem of large-scale multi-users and satisfying the High-reliability scheduling requirements of enterprise managers in emergencies.
The scenario shown above is mainly applied to the monitoring within the station, and it inevitably faces the problem of remote shared access by the superior dispatch center. The traditional streaming media distribution solution should be able to easily achieve multi-level network monitoring. However, if only the softswitch NGN architecture is used, the multicast mode will encounter major problems in the railway industry wide area network and low bandwidth.
Due to the nature of streaming media forwarding, the traditional solution naturally has a multi-tier architecture. At this time, the streaming media server in the station is responsible for the distribution and storage, and is also responsible for the forwarding function. This will inevitably increase the server's performance pressure. However, this solution solves the problem that the railway wide area network cannot implement multicast and has limited bandwidth. It forwards the video images in the station to the upper-level streaming media server, and the upstream streaming server completes the distribution function, which solves the problem of multiple users accessing the same image at the same time.
Due to the multi-level monitoring of railways, the network often cannot enable the multicast function. If the front-end IPC unicast is sent directly to the upper-level platform, it will inevitably consume limited WAN bandwidth, and it is also limited by the IPC's distribution capability. Then, how to make the appropriate modification to the Fig. 2 method so that it can obtain the features of high reliability, large capacity, and high performance, and at the same time, it can adapt to the shared access of the railway industry wide area network?
In fact, the corresponding solution is also very simple, that is, drawing lessons from the experience of traditional programs, adding a media forwarding server to the monitoring platform at the station level can solve the above problem, as shown in Figure 4. However, it must be noted that the monitoring software platform at the station level must be capable of simultaneously supporting front-end video streaming unicast and multicast modes, and can implement automatic unicast/multicast switching functions according to different access requirements. Such a solution can adapt to the large-scale application of high-definition systems in the railway industry in the future, and can ensure the efficient and reliable operation of the entire monitoring system. It is believed that the high-definition monitoring system will surely be more widely used in the railway industry, and how to deal with the impact of high-definition monitoring systems on existing solutions is worthy of everyone's common attention.
Multistage Centrigual Pump,Multistage Centrifugal Pump,Multistage Centrifugal Pump For High Head,Multistage Centrifugal Pump Assembly
Hengshui Yuanhan Trading Co.,Ltd , https://www.yuanhanpump.com