Viasat looked for standards-based technologies and COTS software to build a highly custom teleport bandwidth management system called SAM – Satellite Access Manager. SAM runs on a 2-to-8 core Linux blade server in a high availability mode. They typically do not have the luxury of big footprint hardware. Using Ka-band or Ku-band satellite uplink with ultra small 12-inch tracking antenna, bandwidths reach up to 8-10 Mbps. Even higher data rates are possible with larger ground units. As with every network infrastructure, it needs to be managed and controlled. However, in this environment, very little was truly off the shelf.
To build SAM, Viasat employs a development methodology known as Agile Scrum, which is an iterative and incremental approach to software application development. Small tasks are identified and an estimated commitment for the sprint goal is made, then reviewed, and next tasks are prioritized by customer and internal stakeholder demand. Requirements change and churn based on an unpredictable nature. They accept that problem because features may not be fully understood or defined, so breaking down the tasks into smaller chunks reduces risk and they can respond to customer and market driven deliverables quickly.Another challenge is not knowing what has already been developed and in the public domain. Developers tend to build from the ground up, and keeping aware of the technology stacks available keeps cost and development time down. Core COTS software functionality is out-of-the-box and they can focus on features particular to their space and spend time on actual customer requirements.
The management application has some of your traditional FCAPS functionality, but the main components are to monitor and control equipment and services. Topology is an on-the-move, hub-spoke architecture with a few thousand devices. For fault, it remotely monitors alarms on equipment that has gone down or degrading, equating to loss of service. SAM collects a large number of events and aggregates the data to a consolidated management view. The communication layer between remote and central server is SOAP / XML. Operators have a global view and can detect loss of service proactively before the customer calls come in. Fault correlation reveals very complicated outage scenarios into simple user interface displays. The NOC user is concerned with keeping the lights on, and their job is to simply make sure service is not interrupted. They spend more time on actionable tasks rather than troubleshooting. It's automated through the network management application.
The system handles intense signal processing. Because bandwidth is king, there is a ton of performance trending, business analytics and intelligence data collection. SAM takes in quite of bit of data from Eb/N0 (energy to noise ratios) compared to bit error rate performance to dropped CRCs. Managing device configurations and services are key too. They provision and audit multiple remote users on the network within a common bandwidth pool, provide dynamic assignment of bandwidth and prioritize communications.
SATCOM bandwidth increases and costs coming down prove to be an attractive solution. Provisioning and managing fault and performance data ensures service availability just like any other telecom application.
