As has been highlighted many times, a major focus of attention for vendors like OneAccess continues to be on working on innovations that ensure businesses have reliable high-speed access to their Cloud-based applications.
There can be little doubt that the Cloud is rapidly changing the fundamental nature of computing for businesses large and small. Analysts such as IDC are even going as far as predicting that terms like public and private clouds will eventually disappear from our vocabulary just becoming the de facto standard for business IT provisioning by as soon as 2020.
If IDC is right, the communications’ industry and its supply chain, over the next five years, needs to agree on the standards framework that will ultimately drive the innovation needed to ensure reliable high-speed Cloud access for all businesses and individual users alike. At the moment for some, having a connection that they can depend on can still be a lottery based ultimately on their physical location and local link options.
Application performance and availability are the major factors that are determining the rate of Cloud adoption across the board, with restricted bandwidth and traffic congestion often cited as among the primary reasons for delayed migration. If users cannot be guaranteed that they will not be faced with frequent disruptions and poor quality of experience (QoE) they are unlikely to fully embrace the Cloud in the time-frame that IDC suggests.
In cases where fiber has not yet reached the cabinet (nor is likely to any time soon) the only realistic and viable solution to overcome these objections is to find ways of efficiently aggregating multiple connections to boost the capacity of the available links. There are several multi-path protocols such as IFOM that help boost performance in the Wifi/3GPP mobile networks, but an industry standard approach is yet to fully emerge for the aggregation of Wifi, LTE, xDSL and even broadband satellite links between the CPE and a central hybrid aggregation gateway.
The following diagram depicts the network architecture to deliver an hybrid access by a service provider:
Today’s corporate hybrid access deployments have commonly used session weighted round-robin principles to balance traffic flows across multiple broadband links on a sequential basis, but this relies on each link having the same bandwidth capacity to work effectively, which is not always an available option. Where it has been applied it can offer some improved QoE compared to a single fixed line connection but this rather blunt approach does not seamlessly deal with issues such as link failure or easily allow for dynamic traffic over-flows, that many critical applications depend on.
A more flexible and effective approach is to utilize an alternative packet-based distribution protocol. This enables priority traffic to be routed via the optimum link based on the speed and bandwidth requirements of the application. In practice this allows a Hybrid Access Gateway (HAG) to be set up using multiple combinations of the available bonded links. What’s more, packet-based protocols allow different links to be used as the primary connection, with others used for traffic off-load when the primary is close to maximum capacity or as a fail-over in case of a major outage of one of the connections.
The following diagram describes this packet base solution:
Crucially, packet-based protocols have the flexibility to adapt to the type of traffic, a requirement for applications that are sensitive to delay, performance and QoE
With many companies moving to Cloud and SIP-Trunking for voice communications, as well as the increased use of video-conferencing and other bandwidth hungry applications, the need for intelligence based traffic-flow management that supports dynamic allocation of bandwidth for priority traffic is a no-brainer in my view. Which for me means that as an industry we need to focus our R&D firmly on packet-based distribution protocols as the future for hybrid access deployments.
