Video surveillance, meet wireless mesh: fiber cabling is best for IP security networks, but...

Digital video surveillance is becoming a distributed IP

network application. Closed-circuit TV systems have traditionally been kept siloed, running over duplicate coaxial infrastructures, but the advent of digital video surveillance is enabling the "collapse" of security applications onto primary LANs and WANs alongside mission-critical applications.

The entire video surveillance function is moving to IP, as the old separate networks run by security and facilities management staff evolve to sophisticated technological solutions managed by IT staff. Today's IP technologies allow real-time monitoring of video streams from dozens, even hundreds of high-resolution cameras to a centralized location. Meanwhile, digital technology, improved software and better cameras address the application side.

IP-based video surveillance can improve security and simplify operations. Moves, adds and changes cost less, event correlation across multiple cameras and platforms is achieved using search engines, and standard resolutions for megapixel enable digital zoom capabilities.

Convergence reduces costs associated with build-outs and lifecycle costs of managing and maintaining security networks. The prevalence of wireless and high-throughput Gigabit Ethernet switches makes this transition viable.

The remaining challenge lies in economically aggregating video traffic and collapsing parallel security networks. Fiber-optic cabling and Wi-Fi are both transmission options in this new surveillance network, as is the wireless mesh model, which leverages unlicensed spectrum.

The final hurdle for IT managers is connectivity. Among the concerns are: How best to aggregate and network dispersed cameras? How to ensure high speeds, low latencies and guaranteed quality levels? Can higher-bandwidth video surveillance be migrated from disparate coax networks onto converged IP nets without compromising e-mail, Internet access or voice over IP? Can it be done economically across multiple, geographically dispersed sites ?

Where fiber is in place, aggregation and performance guarantees are expected. Where there is no fiber, the costs and delays associated with obtaining rights and trenching the streets are often prohibitive. Installing cameras where computers and phones are not in use-light poles, garages, delivery docks-may not even be possible with cabling of any kind.

In the absence of fiber, IT managers can consider Wi-Fi, a viable option in sparsely populated areas or scenarios where a few lost frames or seconds is not a problem. Wi-Fi offers low costs of entry, easy build-outs and good performance.

Today, however, "good" may not be good enough. Network administrators should define what good enough means for their network. The quality issue is tri-fold, a function of throughput, latency and interference. Once these thresholds are established, future performance should be projected as networks grow and scalability comes into play.

Wireless mesh technology can negate performance compromises associated with Wi-Fi through ultra-high speeds and the aggregation of hundreds of cameras, ensuring levels of service to multiple applications. Each mesh node in this architecture features Ethernet connections as well as power over Ethernet to operate cameras.

The cameras, dispersed across buildings, are wired to Ethernet switches within, with traffic aggregated from these switches to rooftop wireless mesh nodes. Hard-to-reach spots like lampposts, traffic signals and overhead signs are networked with additional wireless nodes.

This alternative to traditional Wi-Fi is useful in large-scale, centralized digital surveillance networks. High-capacity wireless throughput is sustained while aggregating hundreds of video streams across the mesh topology. The business case improves most dramatically in large multisite deployments encompassing many buildings in campuses, business parks, even entire metro areas.

Traditional Wi-Fi meshes generally provide throughputs per node of 20 to 25 megabits per second--if performance is not subject to interference from other traffic. If it is, that 20 or 25 Mbps can rapidly shrink to 5 or 10 Mbps. The unlicensed mesh alternative is designed to ensure a minimum of 100 Mbps, full-duplex throughput with no latency, and no interference.

While the latency in Wi-Fi mesh networks is tolerable for e-mail and Web access, video suffers with even minimal latency. While typical Wi-Fi deployments incur 10-20 milliseconds per hop delay, the mesh utilizing unlicensed infrared spectrum incurs just 0.1 milliseconds per node.

The Wi-Fi IEEE 802.11b/g standard provides for just three non-overlapping RF channels. With the proliferation of Wi-Fi, interference in dense metro areas is a mathematical certainty. By using the unlicensed infrared band, wireless mesh networks can be deployed without the typical channel contention to provide interference-free communication.

Scalability issues have plagued traditional Wi-Fi networks, as the population of too many access points within a given area causes channel contention, delaying transmissions. With unlicensed infrared spectrum, the reverse occurs. Increased node population improves mesh characteristics and overall network reliability. The new mesh model also addresses the other aspect of scalability--the ability to accommodate whatever happens next.

Fima Vaisman is senior vice president, marketing, for Clearmesh Networks, Pasadena, Calif.

For more information: rsleads.com/706cn-258