Without a doubt, video- surveillance systems are becoming more and more popular. As a result, public-safety agencies are exploring ways to use them in critical locations within their jurisdictions. Besides traditional facility-based solutions—i.e., cameras placed throughout a police department building, jail facility, or city hall—there is a growing interest in placing cameras in remote locations.  Examples of such locations include downtown areas, high-crime neighborhoods, busy intersections, malls, shops, parking lots and garages, parks, and water department sites.

From a system management perspective, it is fairly easy to tie together a multitude of locations into a single solution. Assuming use of a digital, software-based video management system (VMS), and as long as proper Internet Protocol (IP) network connections exist, cameras can be brought together into one unified, virtual system. However, the difficulty concerns the nature of the surveillance-camera traffic.

Traditional IP network applications include Web-browsing and e-mail. In these applications, the traffic is “bursty” in nature, meaning that only periodic bursts of data are sent or received over the network connections; for example, only when the user sends an e-mail or requests Web-site content. After data communications are complete, the network resources are again available for other purposes. Another factor is that these applications are not time critical. Small delays (in the order of less than a second) generally will go unnoticed. In contrast, in the case of a surveillance application, high-capacity video is streaming on a continuous basis, and using up the network resources constantly. Also, video streaming is a real-time application, one in which live video is required to stream without interruptions, in a smooth and predictable manner.

Additionally, with the advent of higher and higher resolution cameras, often referred to as “megapixel” cameras, bandwidth requirements are ever increasing. For example, a traditional 4CIF resolution camera (704 x 480 pixels) would require about 2 Mb/s of bandwidth at 30 frames-per-second (FPS), in high-motion conditions using H.264 encoding. In comparison, the bandwidth requirement for a 1080p high-definition camera (1920 x 1080 pixels) jumps to about 13 Mb/s. (See Figure 1.) These factors make surveillance very much a “killer application” for communication systems, particularly when using wireless networking, which often is the only practical choice for reaching remote sites that are disconnected from the wired network.

A typical citywide digital, IP surveillance network has the following core components: digital IP cameras and edge enclosures; communications infrastructure (wired, wireless and/or cellular); and the back-end solution (including the server, storage, VMS, and monitoring facilities). Figure 2 shows a complete surveillance network with all of the above components, and depicts multiple backhaul options. This article focusses primarily on the backhaul infrastructure design. In order to reach remote locations, when considering wired connections, fiber is generally the only available option (since copper-based Ethernet cabling is limited to 300-foot distances). However, using wireless networking offers the option of building your own wireless network or using a 4G (LTE) cellular connection.

When designing the backhaul to connect remote camera sites with the central management server and monitoring system, it makes sense to first consider the possibility of implementing a wired (fiber) solution. Often, active or dark fiber is available at various government facilities, and sometimes at other locations throughout the city, e.g., in support of traffic applications at certain intersections. Occasionally, pre-deployed conduits are available that make it cost-effective to run fiber to certain locations.

For fiber connections, distance is generally not an issue, since single-mode fiber connections enable distances exceeding 10 miles. The real problem is that extending fiber, without preexisting conduit, is expensive and time consuming. There normally is no business case to justify the high expense of building out fiber if connecting surveillance cameras is the sole purpose. Consequently, wireless technology is the logical alternative. 

Indeed, in most citywide surveillance systems, wireless technology plays a key enabling role for cost-effectively building out the system. However, if the wireless part of the network is not properly designed, the end user is likely to be disappointed with the performance of the network.