Latest News

Raising the Stakes in Network Management

By Anne-Wainscott Sargent | June 1, 2013

      The emergence of Ka-band satellites and ever-more complex terrestrial networks such as 4G are driving greater network capacity gains. With these networks, the number of connected devices and connected components has grown exponentially. Not surprisingly, the stakes for network management – the technology that ensures networks perform as expected and recover quickly in the event of outages or interference – have never been greater.

      “High-throughput satellite networks have greater technical challenges associated with them. They are much more sensitive to atmospheric attenuation,” explains Tim Berdon, vice president of software engineering at GlobeComm, who sees the market increasingly going to high-throughput satellites. “The Ka-band systems have a much higher degree of management requirement, so the control software becomes a key component in operation of the terminal.”

      A decade ago, network management systems were considered ancillary to the communication system, acting as a tool to monitor equipment status. No more. Network management tools today have become “mission critical” whether users are offshore oil and gas operators, banks, broadcasters, or the military.

      “The running expectation in most organizations is that you can get online and when there is a loss of connectivity, being able to quickly recognize it automatically is really important,” says Jim Frey, vice president of research for analyst firm Enterprise Management Associates.

      Frey sees growing demand for network management in virtually every sector, though there has always been an intensive focus on more capable networks in both the communication service provider community and in financial services, where network outages even for milliseconds can means millions of dollars in lost revenue.

      Critical questions network operations people must answer include: “Is current capacity being used optimally?” “Are we actually using the network properly or are we clogging it with things that shouldn’t be on the network to begin with?” and “Are we being smart the way we route and handle traffic loads?”

      Network management tools, both satellite and ground-based, have matured “quite well in the satellite sector,” says Chris Schram, executive vice president of operations for O3b. “Innovations I am most excited about include modem advances that enable satellite signals to adapt and not go down in rainy or degraded environments. Those systems are really coming into the forefront,” he says.

      Schram contends that most network innovation will come from O3b’s customers just like customers have led network advances for GEO constellations. The global start-up operator, which has funding from Google, is set to begin service this spring. Schram says O3b has invested significant time optimizing its network and planning for capacity using leading industry network management tools.

      The first four of O3b’s satellite constellation are set to launch in June, with another four to follow this summer. The constellation, when fully launched, will serve 150 countries. It is designed to orbit 8,000 kilometers above the earth, or one quarter the distance of geosynchronous satellites, making O3b’s satellite performance closer to the speed of a fiber link. As such, the operator can more quickly deploy and coordinate networking and data applications, as well as lay down large data pipes.

      “We have multiple beams per satellite and multiple satellites. As we add more satellites, there is a lot of resilience within the network. It’s a very modular way to build up a satellite network – not gambling everything on one huge satellite,” explains Schram. “We’re dealing with much bigger pipes. If they go down, it’s a huge amount of capacity that will affect a lot of people and potentially end users, so availability, robustness and the ability to clear issues is critical.”

      One technique to keep systems up is diversity switching, observes Globecomm’s Berdon, noting that the company’s broadcast customers frequently build redundant terminals that are geographically diverse that can be brought online if another terminal is affected by environmental conditions.

       

      Going from Reactive to Predictive

      Globecomm’s newest network management system, AxxSys Orion, is about 10 times more efficient in the collection and distribution of data than previous systems. It features an intelligent front-end processor that gathers equipment status from the network edges while decreasing bandwidth use of the software.

      Berdon explains that AxxSys Orion can extend the functionality of the network management system (NMS) without modifying the core software. This allows the NMS to be integrated with business logic systems or to perform advanced analytic functions. These management extensions can look at specific performance values over time in order to perform predictive analysis. In turn, that allows the operator to identify possible equipment failures before they occur. This improves network reliability but it also gives Globecomm the ability to integrate the network management system with other enterprise functions such as allocating bandwidth as orders are placed.

      Ninety percent of our technical support calls are related to network management issues at the remote site. Video streaming is a big challenge because it takes up bandwidth on a continuous basis.

      —Max Usatov, BusinessCom Networks

      “One example would be in the broadcast arena. You can do things like circuit management, which many broadcasters handle manually today. We have the ability to provide plug-ins that trigger the set up and tear down of specific circuits based on rules provided by a scheduling system.” Berdon says.

      Frey sees a major trend in server virtualization, which is causing a ripple effect in the rapidly changing levels of traffic across the network. Application architectures are getting “more and more complex and the growth in cloud services is adding to that complexity,” he says.

      He notes that before, if users wanted to go from five servers to 10, it would have taken a couple of weeks. “With the advent of cloud services, you can ramp up from 10 to 100 servers in a couple of hours.”

      Introducing that amount of new traffic and connected sources of traffic into a network environment results in “many more potential break points,” Frey adds.

      Berdon says predictive analysis is an exciting area as is adaptive networks geared to give maximum throughput in the most cost-effective manner.

      “We’re starting to see networks adapt in the way they deliver the traffic based on what the traffic is, what the specific conditions are, and the traffic priorities. Because of that, management of the network may have to behave differently based on the operational scenario at a given moment in time,” says Berdon.

      O3b is already anticipating this type of requirement with customers like Royal Caribbean, which will have 300 to 500 MHz of capacity going to a single ship once the service launches.

      “Because we are low latency, we are great for voice calls or high speed data, but they also have C- and Ku-band systems on board,” explains Schram, noting that O3b is working with the cruise line and with Harris Corp. on techniques for switching between the three networks, to give them maximum flexibility in running their network.

      Many network management vendors are focused on automation advances. Austin-based Uplogix has a product suite that automatically detects, diagnoses and resolves network-related faults at customers’ remote locations while ensuring connectivity with isolated locations.

      Uplogix currently serves about 40 satellite customers, including Schlumberger, a leading oilfield services provider, which also relies on Uplogix’s network tools to automate routine maintenance operations such as provisioning and configuring devices. The common requirement for all their customers is the need to maintain high service levels across many local remote sites.

      “We’re focused on being able to recover from outages very quickly,” notes James Dollar, vice president of Uplogix. This begins with validating and monitoring network performance much more frequently than prior systems. Validating “where a problem is not is just as important as determining where a problem is,” says Dollar, whose company’s monitoring tool provides reporting every 30 seconds, which is more frequent than the industry norm of every five minutes. “As soon as we find a problem, we triangulate it to the source, and take the first level steps to resolve the problem automatically,” he says.

       

      Automation

      Satellite ISP operator BusinessCom Networks also relies on automation to maintain service levels for its military and government remote users. Since 2003, the Gibraltar-based firm has provided satellite, Internet, fax, point-to-point and private networks for its clients. The company has delivered wireless services to hot spots used for recreational connectivity by a few thousand U.S. Army troops deployed in Iraq and Afghanistan.

      “Ninety percent of our technical support calls are related to network management issues at the remote site. Video streaming is a big challenge because it takes up bandwidth on a continuous basis,” notes Max Usatov, CTO of BusinessCom Networks.

      Usotov points to two sides to the challenge: one, better management at the teleport, and two, better management at the remote site. He’s focused on the second area and his firm’s Unix-based Sentinel bandwidth management server allows BusinessCom clients to prioritize their business traffic such as SAP, Oracle and Citrix, and to ensure that “everyone gets a fair piece of the pie.”

      While automation and proven heritage technology are key components of O3b’s network management platform, Schram cautions that automation can have a downside, too.

      “There can be a high reliance on those devices; but a tool can fail or give a false indication. Sometimes the old-fashioned way of listening, looking at multiple sources of data, and having a good understanding of it helps as much as automation,” he says.

      Toward that end, O3b is investing in the right people in operations and is working with SES, in operating O3b’s NOC, to hire operations staff with multicultural backgrounds and a broader perspective beyond just one discipline so they can bring a big picture view to troubleshooting network issues.

      What new market opportunities are opening up as a result of advances in network management and the growth in high-throughput systems?

       

      SCADA and M2M

      Dollar says as demanding customers move to the more powerful satellite systems, it has opened up Supervisory Control and Data Acquisition (SCADA) networks to use the now-available lower bandwidth networks. The end result: a jump in the number of networks out there – and all demanding reliable connectivity.

      Of course, recent headlines underscore the vulnerability of SCADA networks to security breaches. In November 2011, a water utility in Illinois was reportedly hacked in a cyber attack traced back to Russia. Attackers were able to maintain login credentials and remotely shut down the water pump; the repeated disabling of the pump eventually damaged it.

      Another key market that is opening up for network management advances is M2M, which Berdon says is expected to be one of the largest market opportunities over the next decade. For instance, control systems used to track goods in transit or monitor the refrigerated temperature of perishable foods in containers in transit are just the “tip of the iceberg.”

       

      Integration

      Michael Smith, senior vice president of network technology at Kratos Networks, a division of Kratos Defense & Security Solutions, observes a growing need for bandwidth and end-to-end network visibility by his enterprise and government customers. Increasingly, his software must tie together disparate networks and technologies.

      Wally Martland, Kratos’ vice president of sales, adds that high-throughput satellites require significantly more efficiency to manage the uplink power, since cloudy conditions or a rain storm can affect the transmission up to the satellite. As a result, Kratos has invested in improving its uplink power control algorithms to ensure their systems are even more responsive.

      Smith explains that through continuous monitoring and polling of those devices, Kratos can quickly control and adjust beams based on changing environment needs, capabilities that demand replication and redundancy inside the network.

      So, what should satellite users think about when considering a network management tool? According to Frey companies should first look for solutions that can handle multiple network functions to help them across the lifecycle of the technology they are deploying. Integration and flexibility are key.

      That means understanding and tracking initial deployment, full production, ongoing maintenance as well as being able to support planning and operations, says Frey, including being able to do ongoing monitoring of the network and being able to recognize when changes are being made.

      “The more you can use a suite of products that integrate together in the same platform, the more likely you will be able to stay on top of current status and stay ahead of future problems,” Frey says.

      O3b’s Schram cautions about focusing too much on the best integrated solution. “It really goes back to thinking of the customer first – what are you trying to deliver to them and how do you manage the network best for them? What tools are out there already to do it, and how do you put them together?” Schram adds that it’s equally important to ensure that you suitably match the skills of your people to those tools.

      “If you have a relatively complex network, it’s using tools to make it simpler for the staff you have but also having a level of staff that gets the big picture and works things out. Customers want to call and get somebody who is very qualified and can really look at disparate systems through a variety of tools and find where the issue is and resolve it,” Schram says.