Although video streaming data is also the most common form of data in the network, there are distinct industry features in security. In everyday network data applications, if data packet loss or pause in data transmission is a normal phenomenon, you don't have to be surprised. After we wait patiently, these data will arrive at the designated location in the same number. This is the IP network. The advantages. However, in security monitoring, we have higher requirements for data packet loss and transmission pauses, because we are very particular about real-time applications! We do not want to suddenly get stuck while watching the video, and we do not want some important shots to be lost for no reason at all. Real-time, integrity, and fluency are the basic requirements for security monitoring and video streaming, but it is easy to achieve this. Here are some aspects to analyze ** video streaming related issues:
1, equipment composition:
1): Collector: The network camera is a new generation of traditional cameras and network video technology. In addition to the image capture capabilities of traditional cameras, a digital compression controller and WEB-based operating system are built into the camera. After the video data is compressed and encrypted, it is sent to the end user through the LAN, Internet or wireless network. Network cameras include: high resolution, simplified wiring, intelligent integration to the front-end, intelligent integration, integrated audio, communications security, and strong scalability.
2): Transmission end: The important difference between the network camera and the traditional camera operating environment is that the signal transmission medium is different. The traditional camera transmits through the video cable, twisted pair cable or optical fiber. Generally, each line is transmitted through the “dedicated lineâ€. Bandwidth constraints, and network monitoring, each network camera needs to be transmitted through the network cable or other transmission equipment, and then through the switch access convergence, so it is particularly necessary to pay attention to its code rate (bandwidth occupancy rate), because It is the bottleneck of the network video surveillance system. There are many factors that affect the bit rate of network cameras and network video servers, including: compression method, resolution, frame rate, picture quality, picture complexity, and degree of picture change.
3): The information center side: The resolution of storage and display is also higher. The storage resolution changes from CIF to D1 and gradually shifts to 720P and 1080P** resolutions. Since network cameras usually use dual-stream or multi-stream technologies, the main stream, high-resolution, and high-definition features are used for these services in the video processing services such as near-end storage, video wall display, and video analysis. Great convenience, while remote monitoring uses sub-streaming, which usually has a lower resolution and occupies less transmission bandwidth. As the entire network video surveillance system, mutification is a trend. The MXA image provides a good foundation for system intelligence and provides the most important guarantee for the upgrade of the entire monitoring system.
2, security monitoring system requirements for network bandwidth:
1): Large video traffic: Generally, the D1 stream reaches 2 Mbps, the 720P stream reaches 4 Mbps, and the 1080P stream reaches 6-8 Mbps. According to the 720P dual-stream calculation of each channel, a single code stream needs 4 Mbps per month, and each network channel needs about 8 Mbps code stream. If one code stream is 720P and the other is 1080P code stream, each channel code stream will Greater than 10M network bandwidth. If the switch is 8**, the actual network bandwidth for the uplink port will be greater than or equal to 100M network bandwidth. Considering reservations such as network packets and network switching, you also need to reserve a large enough margin. It seems that ordinary 100M home network switches cannot meet the basic bandwidth. In order to solve the problem of bandwidth caused by the uplink of video data, it is necessary to set up one optical port and one electrical port and two Gigabit uplinks in the replacement of the access layer 8**. Gigabit uplink can solve the long-distance transmission problem of optical fibers, eliminating the need for Gigabit optical transceivers to converge video streams to higher-level switches, while Gigabit electrical ports can connect to NVRs for local storage.
2): Real-time requirements: There is a big difference between downloading video and network monitoring real-time video over the Internet. Generally, when a network downloads video, a large number of caches are locally established. Insufficient network bandwidth, packet loss, and forwarding are not affected in real time. However, for network video surveillance, there is no large cache, and the data needs to be forwarded in real time. This means that under the same network environment, the same video stream can be decoded and played, and the network video can be played smoothly. However, when the video of the real-time network camera is played, serious packet loss occurs. The current switch is basically a store and forward mechanism. The size of the packet data cache affects the forwarding speed. The 8** replacement cache commonly used in the market today is generally 512K. It can still be used in network engineering. When used in network video surveillance, there will be intermittent packet loss (the Caton phenomenon). The actual test shows that there are eight 720P or 1080P cameras, one NVR, and one gigabit fiber uplink switch. If you want to run smoothly, the packet data buffer should be at or above 1M.
3): Convergence capability: Although all of them are 2 Mbps stream, the network camera forwards the maximum network packet at the time of each picture data transmission, that is, each packet size is 1K-2 kbps when each port is Forward with the maximum data aggregation. Forwarding speed, forwarding buffers, and upstream bandwidth will limit network convergence capabilities. For example, when a 2Mbps data stream is forwarded using a 10M network switch, the uplink channel (convergence channel) will take more than 0.2 seconds, and 100M forwarding will take more than 0.02 seconds. If multiple channels are to be forwarded at the same time, the aggregated port needs to be large enough or large enough to cache, otherwise it will delay or lose packets.
It can be seen that it is not easy for **video streaming to be “real-time, complete, and smoothâ€. It is very difficult to achieve the level of security monitoring required for ordinary home-use switches, so when selecting network transmission equipment, Must pay attention to whether this indicator meets the requirements.
1, equipment composition:
1): Collector: The network camera is a new generation of traditional cameras and network video technology. In addition to the image capture capabilities of traditional cameras, a digital compression controller and WEB-based operating system are built into the camera. After the video data is compressed and encrypted, it is sent to the end user through the LAN, Internet or wireless network. Network cameras include: high resolution, simplified wiring, intelligent integration to the front-end, intelligent integration, integrated audio, communications security, and strong scalability.
2): Transmission end: The important difference between the network camera and the traditional camera operating environment is that the signal transmission medium is different. The traditional camera transmits through the video cable, twisted pair cable or optical fiber. Generally, each line is transmitted through the “dedicated lineâ€. Bandwidth constraints, and network monitoring, each network camera needs to be transmitted through the network cable or other transmission equipment, and then through the switch access convergence, so it is particularly necessary to pay attention to its code rate (bandwidth occupancy rate), because It is the bottleneck of the network video surveillance system. There are many factors that affect the bit rate of network cameras and network video servers, including: compression method, resolution, frame rate, picture quality, picture complexity, and degree of picture change.
3): The information center side: The resolution of storage and display is also higher. The storage resolution changes from CIF to D1 and gradually shifts to 720P and 1080P** resolutions. Since network cameras usually use dual-stream or multi-stream technologies, the main stream, high-resolution, and high-definition features are used for these services in the video processing services such as near-end storage, video wall display, and video analysis. Great convenience, while remote monitoring uses sub-streaming, which usually has a lower resolution and occupies less transmission bandwidth. As the entire network video surveillance system, mutification is a trend. The MXA image provides a good foundation for system intelligence and provides the most important guarantee for the upgrade of the entire monitoring system.
2, security monitoring system requirements for network bandwidth:
1): Large video traffic: Generally, the D1 stream reaches 2 Mbps, the 720P stream reaches 4 Mbps, and the 1080P stream reaches 6-8 Mbps. According to the 720P dual-stream calculation of each channel, a single code stream needs 4 Mbps per month, and each network channel needs about 8 Mbps code stream. If one code stream is 720P and the other is 1080P code stream, each channel code stream will Greater than 10M network bandwidth. If the switch is 8**, the actual network bandwidth for the uplink port will be greater than or equal to 100M network bandwidth. Considering reservations such as network packets and network switching, you also need to reserve a large enough margin. It seems that ordinary 100M home network switches cannot meet the basic bandwidth. In order to solve the problem of bandwidth caused by the uplink of video data, it is necessary to set up one optical port and one electrical port and two Gigabit uplinks in the replacement of the access layer 8**. Gigabit uplink can solve the long-distance transmission problem of optical fibers, eliminating the need for Gigabit optical transceivers to converge video streams to higher-level switches, while Gigabit electrical ports can connect to NVRs for local storage.
2): Real-time requirements: There is a big difference between downloading video and network monitoring real-time video over the Internet. Generally, when a network downloads video, a large number of caches are locally established. Insufficient network bandwidth, packet loss, and forwarding are not affected in real time. However, for network video surveillance, there is no large cache, and the data needs to be forwarded in real time. This means that under the same network environment, the same video stream can be decoded and played, and the network video can be played smoothly. However, when the video of the real-time network camera is played, serious packet loss occurs. The current switch is basically a store and forward mechanism. The size of the packet data cache affects the forwarding speed. The 8** replacement cache commonly used in the market today is generally 512K. It can still be used in network engineering. When used in network video surveillance, there will be intermittent packet loss (the Caton phenomenon). The actual test shows that there are eight 720P or 1080P cameras, one NVR, and one gigabit fiber uplink switch. If you want to run smoothly, the packet data buffer should be at or above 1M.
3): Convergence capability: Although all of them are 2 Mbps stream, the network camera forwards the maximum network packet at the time of each picture data transmission, that is, each packet size is 1K-2 kbps when each port is Forward with the maximum data aggregation. Forwarding speed, forwarding buffers, and upstream bandwidth will limit network convergence capabilities. For example, when a 2Mbps data stream is forwarded using a 10M network switch, the uplink channel (convergence channel) will take more than 0.2 seconds, and 100M forwarding will take more than 0.02 seconds. If multiple channels are to be forwarded at the same time, the aggregated port needs to be large enough or large enough to cache, otherwise it will delay or lose packets.
It can be seen that it is not easy for **video streaming to be “real-time, complete, and smoothâ€. It is very difficult to achieve the level of security monitoring required for ordinary home-use switches, so when selecting network transmission equipment, Must pay attention to whether this indicator meets the requirements.
KMP Cat Parts,Cat Engine Parts,KMP Cat Stamping Parts,Custom KMP Cat Parts,CAT Spare Parts
JINING SHANTE SONGZHENG CONSTRUCTION MACHINERY CO.LTD , https://www.stszcmparts.com