"Managing congestion in converged networks"
The company's product and service portfolio is evolving rapidly.
To request up-to-date information, please contact us at info@clear-q.com
The telecommunications industry press is bursting with news of network convergence and talk of triple-play and quad-play: the prospect that most major telephone, cable, and broadband operators will soon be delivering a mix of voice, internet, IPTV and cellular services. But can these operators sustain an experience as consistent and predictable as the plain old telephone service?
The benefits of the Clear-Q technology are most apparent at the edge nodes of any network, which serve a few, or a few dozen, active users, and are most subject to overload. These would be the WiFi access points of a WLAN, the DSLAM and home hubs for fixed broadband access, and the base stations serving cellular data subscribers
The Clear-Q solution preserves a consistent quality of experience for as many users as possible, whether their IP-based services are real-time or interactive. Even when there are more users active than the network is supposed to support, the solution ensures that the overload affects very few of them. And despite its power, the solution is simple to deploy and operate, since it autonomously clears each buffer or queue, with no need to consult policies held elsewhere in the network.
Clear-Q’s mechanism ensures minimal delay for all packets passing through the IP network, first by keeping the number of queues to a minimum, and then by acting as demand approaches overload, before the queue starts to build. This produces a distinctive response to congestion:
1. Short-lived packet flows and low rate media streams like VoIP experience low delay and no packet loss, remaining unaware of any congestion or overload in the network.
2. Higher rate media streams experience low delay and no packet loss, unless they cause overload, when these few streams suffer high or total packet loss.
3. Longer-lived packet flows that employ TCP, the bulk of traditional data traffic, share the IP network capacity fairly with other similar flows.
The Clear-Q technology is a new type of Active Queue Management mechanism, the most widely deployed of which is Random Early Detection (RED). RED is most effective in core routers carrying thousands of active transfers, where its action is statistically significant. However at edge nodes with fewer packet flows the undesirable effects become dominant. In particular, such edge nodes are likely to experience the severe overloads that force RED to revert to simple overflow discard. Given a typical multi-service IP traffic mix, the Clear-Q mechanism maintains its distinctive response to 50% or 100% overload and beyond.
Long-established Internet standards specify how to assure Quality of Service (QoS) in IP networks, but QoS has proved remarkably resistant to actual deployment. In part this delay stems from complexity; not only the effort to develop the policy infrastructure, but also doubt about the operating cost of maintaining all those policies and service level agreements. And narrow QoS solutions solve some technical challenges of packet loss and jitter while degrading the overall quality of experience: multiple queues that store and so delay packets in transit; blocking sessions with Admission Control. But mainly the focus on premium users and services ignores the commercial imperative to deliver consistent quality to all of a network’s customers. The QoS standards appear to take the Marie Antoinette view of the best-effort users. The Queen’s response, when told that her people had no bread, was: “Let them eat cake.”
Clear-Q offers the simple and effective solution to keeping queues clear. And this enables both fixed and wireless IP networks to sustain the profitability of their infrastructure as traffic loads become ever more demanding.
Contact Clear-Q to learn more about how our technology can help prepare your business for the future.