Feature

Overview

With the increasing demand for IP surveillance, it has become the main stream of the surveillance market in recent years. As numerous articles and related news stories have already introduced the advantages of IP surveillance over traditional ones, we are not going to address this again here. However, few articles have discussed network transmission; therefore, we are going to share our experience of this with you.

Currently, the network ports used for IP cameras are generally Fast Ethernet ports. Thus, we can say that the LAN networking of IP surveillance systems is all based on the Ethernet. The major Ethernet technologies, such as VLAN and STP, were previously developed for data communication. Only QoS was developed to transmit voice and video over Ethernet networks.

Because these technologies were developed from the point of view of data communication on an Ethernet LAN, it has not been possible to meet a number of special requirements for IP surveillance networking, such as a rapid back-up link when there is a link failure. The redundant link in RSTP (Rapid Spanning Tree Protocol) can transit from a blocking state to a forwarding state in 1 to 3 seconds, but this is not good enough for professional IP surveillance. Imagine a link failure occurring near the backbone or near a major link in your IP surveillance network; how many IP cameras will lose their video in three seconds? A professional IP surveillance network needs a solution for a super rapid backup link, so that video transmission will not be severely impacted by a link failure.

Hierarchical Network

When planning an enterprise Ethernet LAN, the hierarchical network model is usually considered. The network is segmented into three layers: the access layer, the distribution layer and the core layer, in which each switch has redundant links to other switches for backup purposes.

The Access Layer provides a means of connecting devices to the network and controlling which devices are allowed to communicate on the network. The Distribution Layer aggregates the data received from the access layer switches before it is transmitted to the core layer for routing to its final destination.

The Core Layer is the high-speed backbone of the internetwork. The core area can also connect to Internet resources.

Figure 1 : The hierarchical network model

Ring Topology

Most of the Ethernet swtch manufacturers, such as Cisco and D-Link, design their products based on this philosophy. This network topology usually adopts STP or RSTP as the major fail-over method and includes many redundant links. This kind of network design is good for an enterprise or a campus, but it is not suitable for a large geographic area such as a town or an area of a city.

When a network covers a large geographic area, there will be a high cabling cost; therefore, it is not cost effective to adopt a hierarchical network design. To cover a long distance or a large area with a network, ring topology is most commonly used. Please refer to the illustration below. In such a topology, the Ethernet switches are linked together in a ring. Ring topology saves the cabling cost while retaining the failover function through STP/RSTP or other proprietary ring algorithms.

Figure 2: EtherWAN's α-Ring technology provides a redundancy mechanism to recover from a network failure in less than 15 milliseconds.

Fail-over α-Ring

To provide a rapid fail-over function and solve the large geographical area network cabling issue, we would like to introduce the α-Ring technology of EtherWAN Systems, Inc. α-Ring supports ring topology for Ethernet networking and provides a rapid fail-over recovery time of less than 15ms, making it suitable for IP surveillance network applications. The α-Ring also supports full ring topology as well as multiple rings. Switches can be connected to form a ring, and individual rings can be interconnected to form a larger ring, with redundant backup paths for links between rings. Once there is a link failure somewhere on the α-Ring, the fail-over will take place in less than 15ms. During the link failure and fail-over period, you might not experience any video frame loss.

Figure 3: The α-Ring also supports full ring topology as well as multiple rings.

α-Ring Algorithm

α-Ring technology can quickly react to a link failure because it is specifically designed for rapid fail-over, as opposed to STP/RSTP, which was originally designed for loop free topology and for which fail-over is a derived feature. With STP/RSTP, the switches need to send BPDUs (Bridge Protocol Data Units) continuously for communication. This not only consumes network bandwidth but also consumes the switch's CPU power. Switches that use α-Ring technology only communicate at the beginning of network initiation. After that, there is no communication flow between switches. Each switch only monitors the link status of the α-Ring port, and rapidly reacts to a link failure when it occurs. Here are two flow charts that illustrate the initiation and fail-over algorithm of the α-Ring.

Non-stop transmission is always the first priority, especially in IP surveillance applications. EtherWAN's highly-reliable α-Ring provides a high-speed network redundancy mechanism that can recover from a network disconnection in less than 15 milliseconds, minimizing video frame loss during the recovery process. With the proprietary α-Ring mechanism of setting the last connection as a backup path, the process of ring recovery will not cause any additional data loss when the ring is restored.

Figure 4 : The flow charts of initiation and fail-over algorithm of the α-Ring.