Premium Services Using Vipersat Dynamic SCPC (dscpc) With Guaranteed Bandwidth July 2009
Summary Premium services offered by satellite service providers are far more attractive if Service Level Agreements (SLA) can be arranged through guarantee bandwidth. A common business model is to offer quality of service in the form of a Committed Information Rate (CIR). Customers are willing to pay a premium for service where their access and experience will meet or exceed their expectations. Low latency and low jitter connections are the hallmark of SCPC links. The drawback has always been the bandwidth wasted when the links are lightly loaded, or are offline, and the inability to offer Peak Information Rates (PIR) that exceed the guarantee on a best effort basis. This has been a feature of TDMA solutions, but they do not provide the quality or efficiencies of SCPC circuits. Comtech EF Data s solution is the latest release of the Vipersat Management System (VMS). VMS has long been the industry standard for dynamic assignment of SCPC links. Now, incorporating a guaranteed bandwidth algorithm, it provides operators with a flexible solution for offering the highest quality communications links which can be dynamically sized based on instantaneous usage. Service providers can now announce a service that guarantees bandwidth on a site basis, but can fall to a Minimum Information Rate (MIR) if the site is lightly loaded. If the site is offline, it consumes no bandwidth provisioning the remaining for other sites. However, as soon as there is a need, VMS will resize other carriers (with as little network disruption as possible) to insure the site gets its guarantee. In addition, system operators can provide their users with priority within the dynamic allocation of bandwidth. On a site by site basis, users can be assured that their most important links get the bandwidth they need as soon as they need it. In brief, there are three levels of service: MIR (Minimum Information Rate) A level of service always available to the customer insuring the ability to enter a clear channel SCPC circuit or have a timeslot in STDMA. CIR (Committed Information Rate) Bandwidth that is guaranteed to be available instantaneously based on the users demand or contention. PIR (Peak Information Rate) Bandwidth that can be used by any site on best effort availability, and categorized through multilevel prioritization. These features are thoroughly explained in the following sections below. 2009 Comtech EF Data Page 2 of 8
Vipersat Guaranteed Bandwidth Description Vipersat is the industry standard for SCPC bandwidth-on-demand solutions. Enhancing its innovative dscpc (dynamic SCPC) algorithm to support bandwidth guarantees and site priorities is unparalleled in the industry today. With dscpc, when a remote site modem detects a previously configured application, such as a VoIP call, video conference, or QOS stamped traffic, the modem requests inbound (remote to hub) bandwidth from the Vipersat Management Server (VMS). The VMS then allocates inbound bandwidth for the site, and reconfigures modem(s) and demodulators(s) to support the request. Each remote site can be configured with a minimum rate, a guaranteed SCPC inbound data rate and a maximum SCPC inbound data rate. The minimum rate provides connectivity and service when the link is lightly loaded. However, when needed, the VMS will ensure that a remote can receive at least its guaranteed rate regardless of network utilization. A remote s maximum rate is determined on a site-by-site basis using the site s link budget, the modem configuration options purchased, and the actual service offering. The figures below are from the VMS. The first shown here is the Bandwidth Policy for a site. This is where the Minimum and Maximum bit rates and the site priority are set. 2009 Comtech EF Data Page 3 of 8
The next figure shows the bandwidth reservation page. All the satellites configured in the network are listed here providing the assignment guarantees. The operator can select the current satellite, set the data rate for the reservation and select the Modulation, FEC Rate and FEC Type. An Active Status indicates the VMS is configured to provide the guaranteed rate. If the site is roaming and transitions to another satellite, the VMS will automatically switch to the other satellite providing next level of guarantee. The VMS allocates inbound bandwidth for sites automatically from one or more satellite capacity pools. When a site is using less than its guaranteed rate, the unused bandwidth is available for other sites. In certain cases when providing a remote site with its guaranteed rate, the VMS can move other carriers in the satellite capacity pools, or resize carriers that are currently larger than their site s guaranteed rate. So when it is required the VMS must make decisions about which carriers to reduce. The example scenarios below will explain in more depicted detail. A key objective in the VMS guaranteed bandwidth algorithm is to minimize the movement and/or resizing of existing carriers in the satellite capacity pools when a remote site requests a data rate greater than or equal to its guaranteed rate. 2009 Comtech EF Data Page 4 of 8
Dynamic SCPC (dscpc) with Guaranteed Bandwidth Illustrated The examples below illustrate how the guaranteed bandwidth algorithm operates. The examples were selected for simplicity, but fully demonstrate the concept. Scenario 1: 1) Each site is assigned a CIR of 512 Kbps with a transmission waveform of QPSK ¾, using a 30% allocation factor equaling a bandwidth guarantee, per site, of 443.733 KHz. 2) R1 is configured with the highest priority over R2 - R4 which are equal in priority. 3) Each site is dynamically allocated SCPC bandwidth occupying well above their guarantee. 4) There are two 1MHz bandwidth pools. The first remote (R1) transmits requesting bandwidth that exceeds the size of either pool. Even though the request exceeds the site guarantee the VMS will give it the maximum amount of bandwidth available or equal to the site maximum (an example of moving toward PIR). The image above as illustrated from a VMS display snapshot of the spectrum view, shows that R1 switched to a carrier size of 1.15 Mbps, which fills the pool. When R2 switches into SCPC requesting bandwidth that also exceeds the total available, in this scenario the VMS assigns the same bandwidth as R1 as depicted in the view below. From the spectrum view we can see that all the bandwidth in both pools has been occupied equally by R1 and R2. 2009 Comtech EF Data Page 5 of 8
Therefore when R3 requests SCPC bandwidth the VMS must make a decision as to which of the two carriers to downsize before allocating R3 with bandwidth. Through the reassignment as depicted above R2 and R3 split the bandwidth in the second pool evenly, this is because R1 has a higher priority. VMS deduced not to reduce R1 s bandwidth maintaining quality of service during increased capacity loading. However, when R4 requests bandwidth, we end up with the final solution below giving R4 its CIR as depicted below. The VMS was forced to reduce R1 in order to provide R4 with its guarantee. However, the bandwidth did not split evenly, as it did when R2 was reduced to accommodate R3. This is because R1 is still the dominating priority whereby any excess bandwidth is provisioned over R4 guarantee. Scenario 2: 1) Each site is assigned a CIR of 256 Kbps with a transmission waveform of QPSK ¾, using a 30% allocation factor equaling a bandwidth guarantee, per site, of 221.866 KHz. 2) R1, R3 and R4 have equal priority. R2 s priority is lower. 3) Each site is requesting a peak rate (PIR) of 1Mbps. 4) There is a single 2MHz pool. R1 generates a request for a peak rate of 1Mbps worth of bandwidth. The VMS evaluates the request and sets up the SCPC transmission. 2009 Comtech EF Data Page 6 of 8
R2 then generates a request for a peak rate of 1 Mbps worth of bandwidth. There is enough space to grant the second request so both sites are allocated bandwidth at PIR. Note that the 1 Mbps carriers occupy less than half of the 2 MHz pool as this is a factor of modulation and coding gain. R3 then generates a request for a peak rate of 1 Mbps worth of bandwidth. The VMS considers the following: 1) The amount of contiguous bandwidth remaining in the pool 2) The guaranteed rate configured for R3 (256 Kbps) 3) The maximum amount of bandwidth requested from R3 (1 Mbps) There is enough contiguous bandwidth in the pool to support R3 s guaranteed rate (256 Kbps), but inadequate amount to support the total requested. Because the VMS guaranteed bandwidth algorithm is designed to minimize the impact on existing carriers when providing a remote with its guaranteed rate, in this case, the VMS assigns as much bandwidth (greater than its guarantee) as available to R3 without affecting the existing R1 or R2 carriers. R4 then generates a request for 1Mbps. There is no available capacity, so the algorithm must reduce one of the carriers in the pool. At this point, R2 will be affected as it has lower priority allowing all other sites to take precedence resulting in carrier reduction as depicted below. Note that neither R1 nor R3 were touched as they were of equal priority with R4. Since R2 had to be reduced the VMS provided the site with its guarantee and gave the excess to R4, the higher priority carrier. 2009 Comtech EF Data Page 7 of 8
Satellite Bandwidth Capacity Considerations when using Guaranteed Bandwidth. The VMS guaranteed bandwidth algorithm does not support the oversubscription of CIRs. Therefore, the minimum amount of bandwidth required would be the aggregate of the bandwidth reservations. Excess bandwidth would be available only if remotes were off line, or operating between their MIR and CIR. Bear in mind carrier reductions or movement only occur when the VMS cannot fulfill a bandwidth request involving a site with a bandwidth reservation. To provide a premium experience to your customers a minimum of 10% excess capacity over the aggregate of the bandwidth reservations is recommended. 2009 Comtech EF Data Page 8 of 8