NetGain Communications

As more mission-critical business applications become available over IP networks, there is increasing need for equipment-based restoration/protection processes to keep access links at each location continuously available.

One way to do this is by implementing the Border Gateway Protocol (BGP) version 4 at your router, which balances the traffic between your location and two local IP service providers.

BGP chooses the single best path for reaching either destination based on user-defined metrics. Once the path is selected, all the traffic goes out over that path.

An alternative to per-destination load balancing is per-packet load balancing, whereby individual packets are distributed evenly between the two links.

In either scenario, if one of the links goes down, the other link takes all of the traffic.

Implementing BGP4 to achieve ‘dual homing’ resiliency of your enterprise access to the internet, while a prudent endeavor, is not for the do-it-yourselfer. As with all high-reward efforts, proper configuration and maintenance of the Border Gateway Protocol between yourself and multiple ISPs should be done with planning and the assistance of trusted IT professionals with proven experience.

A construction worker in Baltimore recently drilled through a fiber that knocked out 27 DS-3s. It impacted a few carriers.

In California, a developer doing road work on a new subdivision accidentally tore up 2,000 feet of fiber optic cable, knocking out voice, video and data for 20,000 business and residential customers.

Telephone, Internet and e-mail communications between Anchorage and communities north were significantly disrupted after an Alaska Railroad work crew accidentally severed a fiber link.

Problems like this are fairly common often, especially if you reside in an area undergoing heavy construction. Businesses need to be ready with alternate ways of reaching customers. Distributing private lines among competing carriers, specifying alternate routes, or having telephone numbers that reside on another network can mean the difference between staying open for business or shutting down operations for an indeterminate period.

As users continue to demand instant access to information anytime, anywhere, you may have bumped into some performance problems or slow response times on your network.

Sure, you can add more bandwidth, but that gets expensive. A better way of dealing with the performance problem is by establishing a Quality of Service for each application.

The routers and switches, which control the flow of data on your network, direct data over the best paths. They can also look at a packet of information and see that it is for a phone call or a video conference and give it priority over routine data applications that are not so sensitive to delay, jitter, and packet loss.

By setting QoS you can ensure that your different types of communications will run smoothly. It's all about providing your business with more control over its communications rather than just hoping that everything will work out okay.

MPLS VPNs are increasingly displacing Frame Relay networks, according to Forrester Research.

Forrester found that 48% of businesses surveyed are using VPNs for WAN connectivity between their locations, compared with 23% that use Frame Relay for the same purpose. Indications are that MPLS could become the technology of choice for VPN users.

Why the shift to MPLS? Several reasons. . .

MPLS VPNs provide the means to securely extend the reach of applications, even voice, with the same reliability as the PSTN.

Although Frame Relay will continue for niche applications, it is expensive to add connections, especially if the network must be fully meshed to achieve any-to-any connectivity.

The more economical way to realize a fully meshed network is to use MPLS. Using existing equipment, full mesh connectivity can be achieved with a single connection from each site to the service provider's network. The provider takes care of the routing changes to join all your locations together.

If you have point-to-point connections, DSL and an IP-VPN between locations, you can greatly simplify your network and save money by standardizing on MPLS. A network assessment will reveal whether this is a worthwhile strategy for your organization.

For years industry pundits have been promising that enterprises would one day be able to use a single pipe to the carrier cloud for all of their business traffic, including VoIP. Until then, they had to make do with VoIP trunking between their premises systems.

Although this helped lower toll costs between office locations, these VoIP islands had no IP connectivity to the outside world. Consequently, enterprises still had to maintain separate circuits to the public network.

Today, the single-pipe vision is becoming a reality due to the wide acceptance and implementation of the Session Initiation Protocol (SIP).

SIP trunks are used for call control and routing, enabling enterprises to create a single, pure IP connection to carrier clouds. This is achieved through the use of 'proxies' at each end of the trunk whereby an enterprise SIP proxy peers with a carrier SIP proxy.

Voice is simply layered on top of the network as another IP application, traversing routers, switches, and border elements. SIP sets up and tears down voice calls over this logical trunk. Calls are handed off to a PSTN gateway for the last mile if they are bound for traditional phone connections.

For enterprises, SIP networking means reducing the monthly recurring cost of separate PSTN and data circuits to the premises. This also eliminates unnecessary TDM T1 interfaces on the IP PBX, since hundreds of VoIP calls can be supported through a single T1 interface.

Today, enterprises can leverage SIP for maximum cost savings and network efficiency.

A long held operating principle of the Internet is 'neutrality,' whereby information is passed from sender to receiver, without having to first go through the equivalent of a toll booth. Yet, that's what could happen if Congress goes along with the big telecom companies.

Some accuse the telecom giants of attempting to unfairly profit from their investment in IP networks by charging content providers extra to guarantee prompt delivery of their data. Ultimately all content could be subject to different classes of service based on extra fees for preferential treatment.

Under this scheme, the telecom companies could also favor use of their own search engines, which would steer users to preferred content, or block content from competitors entirely. They could do the same with VoIP services: slow down the performance of competitors' services, while guaranteeing the best possible performance for their own.

There is the very real possibility that the quality and responsiveness of the Internet could suffer, stalling innovation and negatively impacting the economy, which has become 'net dependent.'

Worse yet, Congress will have established a precedent for further meddling with the Internet that could erode its value over time, as it gets shaped and reshaped by political forces.

New entrants into the telecom market are rolling out innovative VoIP services at a furious pace and on a scale that should have incumbent telecom companies worried.

On the consumer side, Vonage spent $414 million in 2005 on advertising to build its brand and expand market share, while cable operators continue to bring voice, data and video to large numbers of residential customers. Now there is a push among small, nimble providers to bring VoIP service to the SMB market.

You would think the incumbent telecom companies would be more aggressive in protecting their SMB base, but they are distracted by the immediate threat to the consumer base and worried about VoIP cannibalizing their SMB revenues.

Ultimately, the incumbents will have to get into SMB VoIP in a big way. They'll reap less margins than traditional voice services, but will be in a better position to sell more services like wireless, reduce churn, and eliminate the cost of replacing customers lost to the competition.

Meanwhile, smaller competitors are leading the charge.

All-in-one security appliances that combine firewall/VPN capabilities with intrusion prevention and virus protection seem all the rage these days, particularly among midsize companies.

Just watch out when you add VoIP to your traffic mix. It turns out that VoIP can experience a 10 - 50 percent performance hit when the traffic is forced through one of these appliances.

Not only are the voice packets being examined more closely, but the box is forced to work harder as it attempts to process lots of smaller voice packets. As you know, delay is the arch enemy of voice, so you don't want to add to this problem.

I'm encouraged by the development of acceleration technology that speeds VoIP processing to ensure the best possible voice quality, despite deeper packet inspection to ferret out viruses. Vendors are also adding QoS to their appliances so that voice gets priority treatment.

Look for these capabilities so you don't have to sacrifice performance for security. The aggravation saved is priceless!

What's not to like about combining the cost savings of VoIP with the added mobility of WiFi?

But reliability is the true measure of this technology. . . Sometimes it works; other times it doesn't. You can save yourself a lot of anguish and user backlash by going with WiFi products that adhere to the 802.11e standard.

Unlike previous WiFi standards that were optimized for high-speed data, "e" understands the nuances of handling voice. Its Quality of Service (QoS) features ensure that VoIP packets are processed with higher priority than routine data, eliminating the chance of packet collisions and retransmissions that disrupt voice conversations.

But QoS is not the whole story. . . the 802.11e standard also includes admission control to limit the number of calls on the wireless network. Both priority packet processing and call admission control provide the reliability users demand.

Just a few things to consider as you transition VoIP calls to your WiFi network. . .