Jack Waters was named CEO of XipLink in 2007 when Montreal-based Xiphos, an aerospace company, looked to spin-off its XipLink acceleration division. Waters was tasked with moving the company from traditional TCP acceleration over satellite toward a broader wireless link optimization system that included advanced stream compression techniques, Internet optimization, VPN acceleration and other value-added functions.

XipLink already was focused on the satellite sector due to kernel-based development of the SCPS standard (Space Communications Protocol Specification) and projects for CSA, PolarSat, MDA, GeoEye, iDirect and others, but XipLink wanted to expand within the satellite and adjacent markets. “The idea was to create a new product culture since, prior to 2007, XipLink was primarily project-based, with specific development requirements on a customer by customer basis. Our push since 2007 has been product-line oriented and market focused.”

Waters discusses his efforts to raise XipLink’s profile and further enhance technology and products. 

Via Satellite: What changes have you seen in the market?

Waters: The customer demand has moved from basic TCP acceleration to a broader protocol optimization system, resulting in a significant market change over the last three years. Fortunately, we knew that was coming by watching wireline and other related markets. XipLink anticipated the market change with proper price points for satellite, algorithms for unique satellite topologies and the right distribution model for satcom. We focused on VSAT first and then added MSS to our plate about a year ago. Most of our competitors began with wireline customers and are now moving into satellite.

Via Satellite: Is it hard for wireline players to enter the satellite sector?

Waters: Satellite is a much more difficult environment than wire-line. When a satellite system is deployed there is more complexity with logistics, personnel, packaging, training and testing than with other markets. For instance, wire-line optimization products typically have many moving parts, take up several RU’s in rack mount configurations or are just physically too large for a mobile satcom deployment, so XipLink offers handheld optimizers, single board implementations or actually embedding our software into co-located modems, routers or remote terminals. One of the key lessons learned in satcom is that ease of deployment is equally important to the product function, thus XipLink launched our Hub Optimizations (XHO) product a year ago, which allowed operators to deploy an appliance at their teleport or data center without having to put an optimization device in the field.

Via Satellite: How do customers influence your offerings?

Waters: The customers expressed their requirements in different ways, mostly by stating they do not typically have trained personnel at remote sites and the cost for travel to a remote site exceeded the benefit of higher throughput. So they greatly influenced us by avoiding optimization technology until a good trade-off could be found, thus, we tested the XHO concept with several key customers, and they loved the idea of using standard browser technology to decompress the traffic even if that is less effective than bracketing the connection with an appliance at the remote end.

Via Satellite: What opportunities like this have you taken advantage of?

Waters: One key event during our spin-off from Xiphos was Mentat’s sale to Packeteer, which was subsequently sold to Blue Coat Systems. They decided not to maintain the SkyX (Mentat) product, which was satellite-specific. XipLink viewed this opportunity to become the de facto accelerator for the satellite world. The second market opportunity was military sponsorship of the SCPS standard. There are two predominant companies in our market that built SCPS acceleration software technology from inception, and XipLink is one of them. The military sponsorship has led to market acceptance of SCPS as an important technology that can withstand the test of time and scale much more broadly than proprietary methods.

Via Satellite: How do you explain to customers the difference between bandwidth acceleration and optimization?

Waters: Acceleration and optimization are now overused terms, but I am not sure who officially declares they are abused. In the satellite sector at the physical (RF) layer, most of the efficiencies have been accomplished with LDPC/DVB-S2 encoding and added techniques like echo cancellation, frequency re-use and other options. Since those benefits are essentially complete, it’s time to move to the network layer for more capacity.

We classify these network layer optimizers in three categories. First, basic (traditional) protocol acceleration uses known techniques such as Fast Start, Acknowledgement Frequency Reduction and a few other known methods to minimize the effects of latency on TCP-IP traffic. Second, in order to optimize TCP/IP protocol traffic over the WAN to a much higher efficiency, companies like XipLink have moved to transparent link optimization by adding advanced TCP streaming compression, Web optimizations, UDP packet optimizations and integrated class-based queuing to achieve the best network layer performance possible without spoofing the application directly.

The third class of optimizers, layer 7 or application optimization, is directly spoofing application protocols to improve performance. Prices are very much aligned along this hierarchy with traditional acceleration essentially free to the VSAT buyer since it is embedded, transparent link optimization at a reasonable capital cost and then application optimization at a significant premium. 

Via Satellite: How do these work together?

Waters: We feel by moving from traditional acceleration to wireless link optimization, the customer typically gets payback in bandwidth efficiency almost immediately, like within a few months at a reasonable capital cost. The additional capacity typically improves application performance considerably. Customers go to layer 7 optimization when the application performance is extremely critical and price is not a factor. Some examples would be financial trading using Citrix when time is the enemy or a large cruise ship generating tens of thousands of dollars in telecom/datacom revenue needs the increased responsiveness to maintain customer attention. 

Via Satellite: Will budget cuts have an impact on your military business?

Waters: There are two aspects that will mitigate the impact of military budget cuts to XipLink — the move to mobility and smaller teams. On the mobility side, our embedded optimizer and lightweight handhelds will replace the rack mountable accelerators at the older but larger field operating bases. Smaller teams mean more devices sold by an appliance and licensing entity like XipLink.

Via Satellite: What advancements have you made in the MSS market?

Waters: We have entered MSS mostly in the BGAN arena, and the key question is how to package and price the solution for this market. Specifically for mobile services, we designed lightweight pocket optimizers that you can easily travel with for land mobile applications. The XipStick is half the size and weight of an iPhone, so the military uses these for both MSS and mobile VSAT applications. The aeronautical business uses XipLink’s XE embedded system to run in pre-certified aircraft routers running on Swift broadband networks to minimize form factor but also to eliminate any FAA-style certifications required for hardware appliances on aircraft. In the Fleet Broadband arena, we have teamed with distribution partners and service providers to address the maritime market with XHO at the distribution point gateway for outbound Web optimization.