Using switches when building a network.

Using switches when building a network.

The problem of ensuring the transmission of signals from video cameras in the network is one of the key ones when building video surveillance systems.

Modern literature describes in sufficient detail and in detail the methods of obtaining a video signal, that is, converting optical signals into electrical signals. Also, specialists are well aware of modern devices that provide processing and storage of information — video recorders and servers.
However, there is still a need to transmit signals from video cameras to devices that provide storage and archiving of video information.
Both analog and digital signals are characterized by the same problems: attenuation (loss of signal power after it has passed a certain distance), noise and interference.
An analog signal is a continuous flow characterized by changes in frequency and amplitude. When a signal fades, its amplitude must be increased. An amplifier increases the overall signal level in the line, including the noise level. Each conversion, each intermediate storage, each transmission over cable or air degrades the analog signal. Eventually, there comes a point when it is no longer possible to amplify, since the noise becomes comparable to the useful signal. It is impossible to regenerate analog signals.
Usually, when transmitting a signal via a coaxial cable, due to the high attenuation coefficient, it is necessary to install repeaters every 50-1000 m. When transmitting a signal via a twisted pair, the following transmitting equipment is used:
— a passive balun installed on the transmitting side with a video signal transmission range of up to 500 m;
— an active device with an output voltage swing of up to 3 V and a transmission range of up to 1000 m;
— an active device with an output voltage swing of up to 18 V and a transmission range of up to 2000 m.
When creating video systems of this type, the signal from the camera is transmitted via a coaxial cable or twisted pair for processing to multiplexers or video recorders, which usually have 8, 16 or 32 inputs. It is easy to calculate that the cost of signal repeaters and wires, especially when creating video surveillance systems for extended objects (for example, industrial enterprises), makes up a significant part of the cost of the entire system.
Digital signals consist of discrete values. A digital signal can only take two values, with some deviations from these values ​​allowed. To transmit digital signals over long distances, active digital devices are used — switches that transmit the signal further with the level of the original signal. That is, it is possible to regenerate the digital signal with each conversion.
Building a network using switches takes it to a new, higher level. After all, to organize backup transmission of a video signal in a direction, which should be taken into account in highly reliable video surveillance systems, it is necessary to split the video signal, use video stream switches, and calculate the video signal transmission system in such a way as to minimize data loss when cable lines break.
Switches not only improve the quality of the signal, but also collect the signal from several sources and ensure its transmission over a certain distance via one cable, and not via a significant number of wires, as in the previous example.
At the same time, the switch does not create network congestion. It transmits information from one segment to another only if such information is necessary, which increases the overall performance of data transmission in the network and reduces the possibility of unauthorized access to data. In addition, the switch is a router, selecting the data transmission route.
At the same time, switches support both traditional Ethernet protocols and their own.
Currently, both standard switches: CISCO, D-LINK, Allied Telesyn, NORTON, ALCATEL, Edge-Core, and industrial switches: HIRSCHMANN, RuggedCom, CTRLink, Korenix, N-Tron, Taiko Network, SixNet, Garrettcom, etc. are widely known and used.
Let's consider in detail the principles of signal transmission using the example of individual switches of each class.
All modern switches provide routing and backup using the standard Spanning Tree technology. This technology, defined by the IEEE 802.1d standard, selects the most rational route for data transmission. Switches operating using the Spanning Tree algorithm automatically create a tree-like configuration of connections without loops in a computer network. This configuration is called a spanning tree — Spanning Tree. The spanning tree configuration is built between switches automatically using the exchange of service packets. This protocol is organized as follows. The creation of a spanning tree begins with the selection of a root switch, from which the tree will be built. The switch with the lowest identifier value is automatically assigned as the root bridge. But this choice may not be the most rational. Then the network administrator, based on the network structure, can manually assign the lowest identifier to any switch. This device will be the root bridge. Then the root port is selected for each of the other switches in the network. The root port of a switch is the port that has the shortest distance to the root switch on the network. Then, designated ports are determined. Each segment in a switched network has one designated port. The designated port of a segment has the shortest distance to the root bridge, among all ports connected to that segment.
When relaying frames, each switch remembers the minimum distance to the root for each of its ports. When the spanning tree configuration procedure is complete, each switch finds its root port — this is the port that is closest to the root of the tree.
That is, Spanning Tree technology provides a search for the most rational path from one point of the network to another. A classic local network is built on a tree structure. Lower-level switches are connected to the root switch, and even lower-level switches are connected to them, and so on. When building a complex local network, a structure with redundant connections between network devices is used instead of a classic tree scheme to increase reliability. Such a network is usually built on switches that support the IEEE 802.1d standard (Spanning Tree — spanning tree). They allow you to create a logical tree (structure) of the network and avoid logical loops.

Fig. 1 shows the connection diagram of Nortel switches

The IEEE 802.1p standard allows traffic to be divided by importance and frames specifically marked as important to be sent first. This technology makes it possible to transmit sound or video during live broadcasting without interruptions.
The IEEE 802.1s (Multiple Spanning Tree) standard allows logical trees to be built for several virtual subnets of a large local area network.
Using switches that support this standard makes it possible to increase the fault tolerance of the network, as well as balance it.
The IEEE 802.1Q standard (VLAN — virtual local area network) allows you to build independent virtual (logical) networks within one physical network. Data in virtual networks circulate independently and do not penetrate from one network to another.
However, Spanning Tree technology also has a drawback: if there are more than seven switches in the network, it may take up to several minutes to restore communication in order to detect and bypass a communication line failure. During this time, all network solutions will be isolated. Of course, this is unacceptable when building security systems.
In such cases, it is advisable to use the IEEE 802 1.w standard. In case of network failures or broken connections, the Rapid Spanning Tree Protocol (IEEE 802.1w) is used, restoring operability along backup routes in a fraction of a second. The Spanning Tree Protocol was developed quite a long time ago, in 1983. Since then, it has been improved. Thus, to eliminate the limitations of STP, which interfered with some routing functions of layer 3 switches, the IEEE 802.1w Rapid Spanning Tree Protocol (RSTP) was developed. A significant difference between the STP 802.1d and RSTP 802.1w protocols is the method of switching ports to the forwarding state.
In HIRSCHMANN switches from Hirschmann Elektronics (Germany), the Hiper Ring ring backup technology is additionally very successfully implemented. This technology ensures continuation of signal transmission in the other direction if the network is damaged at any point (Fig. 2.) At the same time, the time of full restoration without loss of information is <500 ms, one ring can include up to 50 switches, the total length of the ring can be up to 4000 km when working with both Fast Ethernet and Gigabit Ethernet.
The HIPER-Ring technology is based on the concept of backup connections, and it significantly exceeds the «office» Spanning Tree technology. When implementing HIPER-Ring technology, switches are connected to each other, forming a ring in which one of the connections is backup. The control switch sends out test packets and checks the network for serviceability. If a failure is detected, it activates the backup connection and sends data along it without losing information.

Fig. 2 Connection of Hirschmann switches using the Hiper-Ring ring technology.

The time it takes to restore network functionality when a communication channel is interrupted, simulated by disconnecting a patch cord, when using the Hiper-Ring technology is such that changes in the transmission of a digital television video signal are usually not recorded visually or in the archive.
To explain the principles of organizing the Hiper-Ring technology, it is necessary to dwell on the following technical features.
Since Ethernet is a bus architecture, if a ring or loop is formed, any Ethernet broadcast frame will be sent around the loop, causing a broadcast storm and bringing the network to a standstill. However, HIPER-Ring takes this limitation into account. In addition to the standard functions, switches based on this technology can create a physical ring by connecting both ends of a traditional Ethernet bus. Although the Ethernet bus is physically closed, the switch logically breaks this bus. Therefore, transmitted frames will not be looped.
Logically, such switches have two sides (the connection between them is the backup link), each of which continuously transmits diagnostic messages to the other side and receives them in real time in the ring. It should be noted that network failure messages are assigned high priority according to the 802.1p standard.
As is known, building networks based on switches allows using priority traffic processing, regardless of the network technology, if the switches support the IEEE 802.1p protocol. In this case, the switches buffer frames before sending them to another port. The switch usually maintains several queues with their own processing priority for each input and output port. As a result, the switch can be configured, for example, to transmit one high-priority packet while transmitting several low-priority packets.
High priority allows network failure messages to pass through the fastest path through any switch in the ring that supports the 802.1p standard. This ensures that you receive a real-time report of the actual state of the network at any given moment.
In the event of a ring failure, i.e. when a node or cable fails, the switch will still transmit to both ring ports, however, due to the failure, not all devices in the ring will receive diagnostic messages. In this case, both sides interpret this loss of diagnostic data as a network failure. When a failure is detected, the switch activates an internal link, connecting both sides, which returns the network to a fully operational state. The detection of the failure and the process of «healing» the network will take on average from 20 to 300 ms, depending on the size of the ring.
In addition, the system itself will determine the location of the failure and immediately send all information about it to the service personnel. Localization of the failure will now take much less time, which means that the time to fix it is also reduced.
When using HIRSCHMANN switches, it is possible to use Dual Homing technology – redundancy of connections (Fig. 4), which provides a recovery time of less than 3 seconds, redundant connection of network segments to the backbone, detection of failure of equipment to which a backup connection is made, detection of failure (break) of cable systems of communication channels or the covered section of the network.

Fig.4 Connecting Hirtschmann switches using Hiper-Ring technology (Dual Homing)

Switches from Taiwanese company Edge-Core can be 24/48-port 10/100/1000 Gigabit Ethernet autonomous managed switches. These switches implement not only IEEE802.1D (Spanning Tree) protocol, but also Edge-Core's proprietary Fast Mode Spanning Tree, IEEE 802.1s* (Multiple Spanning Tree) and IEEE 802.1w* (Rapid Spanning Tree) protocols, and also support the following protocols:
IEEE 802.1s Multiple Spanning Tree
IEEE 802.1w Rapid Spanning Tree
Supports Accton's proprietary virtual network management functions in Fast Forwarding mode based on ports. It also has an input for a backup power supply on the rear panel to ensure power supply fault tolerance, and also operates at temperatures from -40C to +70C.
The connection diagram of these switches is shown in Fig. 5.

Fig. 5. Scheme of construction of a local network based on Edge-Core switches.

As can be seen from the description and from the table, the ring redundancy protocols for switches from different manufacturers may differ from each other.
The redundancy technologies eRSTP (Ruggedcom) and S-Ring (Garrettcom) are capable of restoring a «ring» network in less than one second, even if the devices are hundreds of kilometers apart.
Korenix has found a solution to this problem in its JetNet 4500 series industrial managed switches. The JetNet 4500 series switches allow you to connect switches with different connection redundancy protocols into a single network and redundantly perform the resulting chain as a whole using Super Ring technology. The essence of the solution is the ability to connect switches with different connection redundancy protocols (such as IEEE 802.1w Rapid Spanning Tree Protocol and Super Ring), including the Hiper-Ring protocol from Hirschmann, into a single network and redundantly perform the resulting network as a whole. Redundancy is achieved by connecting two ports of two JetNet 4500 switches that support Super Ring with two ports of a commercial switch that uses the RSTP protocol. In this case, one of the ports is active and is used for data exchange, while the other is in a waiting state. The waiting port regularly monitors the status of the active port and, if the connection on it is lost, in a matter of seconds «intercepts» control. At the same time, the process of switching to a backup channel remains «transparent» for end users. Fig. 6 shows examples of building R.S.R and Dual Homming networks.

Fig. 6 Examples of building networks using Korenix switches using R.S.R and Dual Homming redundancy protocols.

The software for SixNet industrial switches supports RSTP redundancy technology, QoS for setting priorities, VLANs for network delimitation, IGMP for traffic filtering, etc.
These are the advantages and opportunities that the use of switches provides installers when broadcasting digital video signals.