The principle of low power consumption and high energy efficiency in video integrated control system

As video surveillance systems move towards massive resource distribution and social applications, the technological development trends of IP, high definition, and intelligence have brought the video surveillance industry into a new period of transformation. In the process of large-scale transformation and upgrading of traditional analog video surveillance systems, and in the process of designing and selecting new video surveillance system architectures, users have not only put forward increasingly stringent requirements for image clarity, real-time performance, reliability, and ease of use, but also put forward comprehensive system considerations for energy saving, high efficiency, and environmental protection for equipment manufacturers and system integrators. If two products with the same characteristics and prices are given, users will definitely choose the product with the lowest energy consumption in order to achieve green environmental protection.

A complete video surveillance system includes front-end acquisition, intermediate transmission, back-end display management and storage. The equipment responsible for back-end display management is the core control center of the whole system. Whether it is a single system formed by one node or a networked system composed of multiple nodes, whether it is analog video access, digital video access or analog-digital hybrid access, only through the back-end display management equipment can the dispersed image resources be aggregated and processed and a human-machine interface be formed. The core role of this video integrated centralized control makes it an indispensable soul in every system. Due to the complexity and diversity of this system architecture, especially the need to solve the problem of integrated access of analog video and digital video, the specific composition is also the most abundant, involving receiving optical terminals, network video decoders, matrix switching controllers, hard disk recorders, management servers, disk arrays and other equipment. Since the energy loss of the camera/. responsible for front-end acquisition and the cable optical fiber responsible for intermediate transmission is relatively low, the goal of reducing power consumption, improving energy efficiency and compressing comprehensive costs is placed on the back-end display management link. There are several common problems with this type of equipment:

1. Low integration and troublesome deployment

In the current analog video access systems that have a large proportion in the market, most of the back-end display management devices are equipped with matrices for overall quick browsing, real-time non-compressed display and instant control of all videos. A matrix with 256 inputs and 32 outputs usually requires more than two large card-type chassis, and an external CPU controller and code distributor, etc., not including the hard disk recorder/video distributor, etc., which already occupies nearly the space of a standard cabinet. If the video capacity exceeds that of a single chassis, it requires the coordinated work of multiple CPU units between different chassis, and the connection of a large number of video cables between multiple chassis, which is easy to introduce signal interference and unstable factors. If an embedded DVR recording plus digital decoding matrix or a network video codec solution is adopted, a large amount of network resources need to be configured, the reliability/stability of the system is easily affected, and the cost of operation and maintenance is high.

2. Capacity expansion is limited and evolution is difficult

In the current video surveillance system, the diversity of various video formats such as analog standard definition CVBS, analog high-definition SDI, digital standard definition D1, and digital high-definition 720P brings uncertainty to the evolution of the system and high equipment upgrade costs to users. In addition to the two types of analog video signals that belong to international universal standards and can be directly converted, digital video signals currently have their own compression algorithms and standards for each manufacturer, which are difficult to be compatible with each other, resulting in equipment replacement and abandonment, occupying storage space and polluting the environment. Even the analog matrix responsible for video switching control has structural design limitations. For example, each functional board can only be inserted in a fixed slot on the backplane, and cannot be flexibly combined. In addition, except for the video input and output modules, it is impossible to add new functional units for system expansion.

3. High power consumption and high emissions

In the past, power consumption was often given a lower priority than most other variables (speed/performance, cost, time to market, system risk). However, in today's market, power consumption has become a very important component in the design decision-making process. For example, of the hundreds of millions of yuan in electricity bills spent on various monitoring centers each year, 25% of the costs are directly attributed to the power supply of large equipment and various servers in the system, while 50% of the costs are used to power cooling systems and fans to eliminate the heat generated by running equipment. In addition, even a small monitoring center will emit 20 tons of carbon dioxide each year due to the operation of equipment.

The principle of low power consumption and high energy efficiency in video integrated control system

1. High-density and compact design

Reducing the number and size of control and management equipment can not only reduce system power consumption, but also reduce floor space. For example, the hard disk recorder has evolved from the early industrial control DVR to the later embedded DVR, from installing compression cards to 16-channel CIF and even the current 16-channel D1; for example, the hard disk capacity has increased from 80G three years ago to the current 1T, which has become the mainstream; for example, the matrix switching controller has evolved from the early single-chassis 192 inputs and 16 outputs to the current mainstream single-chassis 256 inputs and 64 outputs in the industry, and the industry's most compact small system with 32 inputs and 16 outputs in a 1U chassis for vehicle/conference room applications; for example, the optical terminal has evolved from point-to-point transmission to single-fiber transmission of 16 channels, as well as node-type/convergent optical systems.

The high-density and compact structure design allows the same space to accommodate more video access, provide better image quality with the same power consumption, occupy less bandwidth resources and disk capacity with the same video indicators, and the same system equipment costs less. For example, a large-scale intelligent network matrix with 256 inputs and 64 outputs, the weight of a single video input and output module does not exceed 1kg, the weight of the matrix with full configuration does not exceed 25kg, and the full load power consumption of the whole machine is only 90w, which is 70% lighter and 85% less than the previous generation of analog matrix. In addition, since the CPU unit/code distributor/power module are all built-in, the power supply demand of supporting equipment is reduced, and the high integration and high integration also greatly improve the speed of system installation and deployment, without considering the interconnection and repeated settings between multiple chassis and multiple devices. The high-density 7U rack-mounted chassis reduces the space requirements for the computer room/cabinet and saves user resources (as shown in Figure 1).

2. Modular structure configuration, flexible expansion

The video surveillance system is divided into multiple functional modules according to different customer needs and technical implementation methods, including: audio and video switching, front-end cloud mirror control, alarm collection linkage, digital video storage, network remote transmission and data transmission forwarding, etc. After years of evolution and development, each functional module has become its own independent industry and product line. Each device realizes different functions and plays different roles to meet the relatively independent needs of users. Therefore, it is designed into different functional boards and hardware structures, and the size specifications and appearance processes of the chassis are also different. However, in the environment of increasingly fierce all-round competition and the general trend of focusing on industrial solutions, it is necessary to consider how to integrate multiple different categories of products to provide users with integrated products and systems with high integration, easy use, and efficient operation and maintenance; and in order to cope with price competition, it is necessary to achieve the industrialization requirements of cost compression and quality control. Even the products of the same company face the problems of low efficiency due to the large variety of products and low output, different hardware structures, non-universal boards and chassis, and inability to guarantee inspection standards. Modular structural design has become the most effective solution.

Taking the current more advanced MSIP universal module structure technology as an example, its core design idea is to unify the hardware interface and underlying protocol stack of various functional modules such as video input and output, audio input and output, Chinese character overlay, full cross switching management, optical reception and transmission, alarm input and output, network video encoding and decoding according to the standard protocol, and assemble them on the universal high-speed bus backplane through the embedded guide rail card slot of the standard rack-mounted chassis. Various functional modules are coded according to the unified standard attribute category, and can be flexibly matched according to actual needs, built into any slot of the universal bus backplane, and the number and location of the board cards do not need to be limited by the address. The system automatically scans and identifies them and quickly goes online. In the future, customers can expand, upgrade and maintain various functional modules by themselves, and the system deployment cost is reduced by at least 20%. For example, the VMS video integrated control platform system developed based on this platform adopts a 7U standard plug-in chassis with built-in CPU unit, code divider, network switch and power module. It supports 256 analog video inputs and 64 analog video outputs, multiple network video codec modules, digital video input and output, and mixed switching control of analog and digital video signals. Users can remotely browse and control the front-end image scene at any node in the network. It can also have built-in multiple 32-channel alarm signal input modules or network alarm hosts to support various alarm types of perimeter alarm detectors (as shown in Figure 2).

Analog video and digital video can achieve unified specifications by using codec modules with the same interface standards, plug and play, and solve the compatibility problem of multiple video formats. The software unit for comprehensive management of video images also needs modular design. Users can configure CMS general control management, video storage, user permissions, proxy forwarding, WEB services, alarm management, GIS maps and other software units in a menu-based manner according to the actual needs of the system. They can be deployed independently or work together.

The main function of multi-module integration is to achieve comprehensive linkage with sound and image, such as associating preset one or more video switching display and video storage after the perimeter alarm detector triggers an alarm, associating preset video switching display after the audio signal exceeds the set decibel threshold, associating preset video switching display after the video scene senses that there is an object moving in the set area, and if the front-end camera is a pan-tilt zoom camera, it will start the preset position and cruise track. For this kind of comprehensive linkage intelligent preset processing, macro instructions can be used to achieve quick editing and custom call. Macro is a user-defined operation instruction that replaces a series of time-consuming and difficult to remember repetitive keyboard operations manually, automatically completes various preset operations, and provides emergency response plan processing for emergencies. Through the friendly human-computer interaction interface of macro, the system can automatically realize the unified call and associated operation of each extended function module by inputting the requirements, and the user does not need to care about how the underlying hardware devices realize command intercommunication and data exchange. The root cause is that all functional modules adopt a unified protocol stack and standard interface design to form a highly intelligent integrated device.

3. Unified platform application, smooth evolution

Faced with a large-scale video surveillance system with numerous and complex devices, the most inefficient management link is the inability to operate quickly and effectively. To this end, it is necessary to start from two aspects: unifying the platform application interface and strengthening the event flow management of a single product.

The unified platform application interface requires that all software units and hardware modules be managed through a master control server, with unified data exchange, unified clock, unified video transmission, and shared processing resources. Taking the VMS video integrated control platform system as an example, this master control server can manage all the basic function modules and extended function modules in the system, including all settings for video input and output, storage disk, recording channel, user authority, event triggering, and operation log. Users only need to access this IP through the client, which not only saves network resources but also improves execution efficiency.

Event flow management pushes the single-core hub to a multi-core node. The core idea is to decompose the relatively large video integrated control platform system into secondary and tertiary sub-devices, so as to facilitate the low-cost and rapid deployment of small and medium-sized systems. For example, the intelligent network matrix uses WEB centralized control technology, with audio and video information flow as the main data reference line, to bind network video input and output, alarm input linkage, alarm partition control, user authority management, front-end operation level and other functions, and integrate multiple business modules. Users can manage conveniently through IE browsers, without the need for dedicated workstation servers, complex connections and tedious debugging, and network keyboards based on standard protocol interfaces also provide users with a convenient and flexible human-computer interaction interface.

4. Low power consumption, high energy efficiency

Whether it is a high-density compact structure design or a modular interface universal high-speed bus, it is necessary to consider the use of a variety of innovative technical means in system design and product design, and use chip processing solutions with higher main frequency, higher performance, and smaller package volume. In many complex system designs, FPGA is a good choice, which can help designers improve the system's ease of use and scalability, and improve unit density. For example, in the video switching and character overlay circuits, the original general solution needs to configure multiple chips, which makes the circuit complex, increases the PCB circuit board area, and reduces the system integration. Through FPGA, the video format can be automatically identified and a synchronization signal can be generated to achieve video synchronization and jitter-free switching. At the same time, user-defined character graphics can be superimposed on multiple video signals at the same time, and the execution efficiency and energy consumption compression rate are increased by 8 times. For example, the latest S series 32-input and 16-output video matrix, the chassis is only 1U high, and it adopts multi-core single-board platform technology. Switching, control, exchange, and overlay are all solved by a main chip, providing the best choice for compact small systems.

Power consumption is a large comprehensive cost expenditure, because when dealing with the heat problem caused by excessive power consumption, the complexity of circuit board design increases, the requirements for port density and bandwidth increase, but the waveform factor decreases, forcing development engineers to adjust project schedules and budgets. (Figure 3)

Chip energy consumption includes many aspects, such as FPGA power consumption comes from pre-programmed static device power consumption, surge programming current, programmed static power consumption, and dynamic power consumption. To solve this problem, on the one hand, using smaller chip process technology such as 65nm can solve these problems, and on the other hand, deeply tap the energy-saving potential and use a variety of power-saving technologies to reduce the power consumption of the whole machine, such as using low-power and high-efficiency DSP and PCB, using intelligent software power-saving technology, and power control technology (as shown in Figure 4).

Optimize the power supply design, reduce the number of components, reduce the PCB area, reduce the temperature, and improve system reliability. Through the power supply intelligent temperature control technology, the rectifier module is automatically turned off/on, so that the power supply clock works in the most efficient load rate range (as shown in Figure 5).

For the chassis structure, the thermal management system is simplified, and a smaller (or no) heat sink is used to reduce airflow requirements. On the one hand, natural heat dissipation can increase the thermal capacity by 20% under the same volume. On the other hand, the heat dissipation system can automatically adjust according to temperature changes to reduce noise and power consumption (as shown in Figure 6).

Low-power video: the future of integrated control systems

Energy conservation, emission reduction and low power consumption design may still be a new concept in the video surveillance industry. The video integrated control system based on multi-standard video access solutions is a brand new platform. Analog video and digital video have their own market share in current applications, but from the perspective of technological evolution and development, it is an irreversible trend from analog to digital. In the process of video surveillance evolution, the access rate is constantly increasing, the access resources are constantly enriched, and the access technology is constantly developing. The current video surveillance system is a system where multiple video standards coexist and needs to be integrated and evolved. The inevitable trend of technological development means that traditional equipment will need to be updated and replaced due to continuous aging. Can it access multiple videos to achieve flexible deployment and guarantee initial investment? Can it respond to the rapidly changing upgrade and expansion needs to achieve low power consumption and high energy efficiency operation? So far, after years of research and development, Tiandi Weiye has mastered several core technologies for switching display and control forwarding in video integrated control management, and will continue to invest in the optimization and energy saving research of the next generation system construction mode. No matter what problems the future video surveillance system faces, reducing power consumption will be a long-term challenge.