Packet loss is a common problem of data transmission in IP networks. Sometimes, packets of data fail to reach the required destination. Whatever the cause – and there are many – if packets don’t arrive at their destination, then problems can result. For some applications, this doesn’t matter too much, provided the rate of packet loss doesn’t exceed a certain threshold. But other, more critical applications, depend on the receipt of all transmitted packets in order to function correctly.

 This doesn’t just apply to end-to-end applications, such as conversational voice or video conferencing, but can also affect those applications that are designed to monitor the performance of the network. The impact of packet loss on such systems is also variable. For example, solutions that are deployed to gather network analytic data may not need to gather all packets that are transmitted. It can be sufficient to sample packets and to extrapolate from averages or samples in terms of packet use and demand.

 However, applications that are designed to protect the network from invasive attack must be able to reliably collect all packets transmitted in order not to miss those that might reveal signs of attack or performance degradation. Worse, some forms of network incursion may effect a rapid escalation of packet loss, so the early signs of such attacks need to be caught before real harm results.

 That’s why its essential to ensure that the solutions that protect our networks from harm are able to effectively capture all packets, with zero loss – and to ensure the resulting data can be screened by the intelligent applications that are designed to ensure optimal and secure network performance.

 It’s possible to develop such solutions using off-the-shelf network interface cards and high-performance CPU hosts. But as network transmission rates increase from 1Gb/s to 10Gb/s and beyond, it’s difficult for such solutions to deliver the performance guarantees required to ensure network protection. The cost of deploying such systems also escalates dramatically, both in terms of CAPEX and OPEX through power consumption and real-estate demands. And, such investments may be at risk through performance degradation.

 What’s needed are specialized solutions that enable network security applications to collect all required packets in real-time and at wire-speed without placing additional demands on CPU performance. By sharing the filtering and processing demands between specialized line interface cards and the CPU, overall system CAPEX can be reduced. Moreover, such solutions can also be deployed in small-footprint chassis units with lower power demands, reducing OPEX and providing the scale required to enable network-wide security solutions.

 There are a number of applications that have strict requirements in terms of packet loss, those associated with the emerging field of cyber security among them. By deploying hybrid solutions consisting of high-performance packet-processing interface cards and CPU hosts, not only will you not lose a packet, you won’t spend one either.

Why not find out more about our high-performance packet processing solutions or get in touch with our experts to see how we can help? Or, come and see us at booth 4354 at CTIA Wireless!

Telesoft Technologies is pleased to announce our involvement in the CTIA Wireless event in New Orleans from May 8^th – May 10^th where we will be demonstrating our industry-leading range of 10, 40 and 100Gb capture cards.

CTIA Wireless is the main event for the global wireless marketplace, attracting more than 40,000 attendees every year with top-level speakers from across the industry. It’s a valuable opportunity for Telesoft to meet with our partners, discuss their requirements and demonstrate our latest technology developments.

This year, Telesoft will be exhibiting at booth 4354 in the Emerging Technologies zone where we will be showcasing the very latest developments in high performance packet capture technology.

Intelligent 10, 40 and 100Gb capture cards for monitoring and analysis 

Designed for high-speed packet capture, analysis and reporting, our high-performance platforms enable 100% accurate packet capture at all line-rates up to and including 100Gb/s.

Our latest hardware platforms use hardware acceleration to deliver zero packet loss and “apply on the fly” real-time configurable filters to ensure that DPI applications can effortlessly connect to 1, 10, 40 and 100Gb networks in scalable, compact platforms.

Contact us today to book a meeting with one of our consultants. We can discuss your needs and apply our expertise to solving your network challenges.

A recent article has highlighted some of the problems that Mobile Network Operators (MNOs) face when attempting to manage traffic in their network.

Traffic management is needed for a variety of reasons. First, quality of service needs to be maintained – both for the network overall and for individual customers. Secondly, with MNOs increasingly offering packages that set limits to data consumption, they need to be able to monitor usage of individual users to ensure that they do not exceed a given quota or that they are billed appropriately if they do so.

Thirdly, there may be restrictions on the use of certain applications, because the MNO or regulator has a specific policy or because there is a legal reason preventing access. Fourthly, there may be specific policies that need to be enforced – for example, if users pay for a high-performance service, the MNO must be able to deliver the required level of performance to the subscribers. And, fifthly, users need fair access to bandwidth – those that over-consume may be subject to temporary restrictions, ensuring that bandwidth is available for other users. We could go on, but it’s clear that traffic management is absolutely critical to the smooth operation and performance of the network.

And that’s why MNOs need the tools to ensure that they can review traffic and manage network demands in real-time. Deep packet inspection technology is critical to this. And with transmission rates increasing all the time, it’s a struggle to keep up. MNOs need DPI solutions that can scale effectively and cope with continually increasing traffic flows.

The volume of data and the rate at which it is transmitted just keep on growing. Traditional DPI equipment will be overwhelmed by this flow of data, as it relies in CPU-based technology for processing the data. A better solution is to perform the pre-processing at the level of hardware, ensuring that only packets of interest are sent to the control application. This means that the pressure on CPUs is alleviated, ensuring the DPI application can perform as expected.

After all, DPI is now fundamental to ensure network performance, to meet customer expectations and to ensure fair access to data for all. Quite simply, without DPI, MNOs would be unable to manage their networks.