Smart Redundancy

What is the problem?

A traditional hot-standby redundancy scheme includes a main (active or online) piece of equipment and a similar piece of equipment in standby which can momentarily replace the main equipment, should the latter develop a fault. Having additional equipment in standby increases the overall reliability and availability of the Hub. The additional investment in reliability is paid back through higher quality of service and enhanced customer experience.

However, such redundancy appears to be excessive when applied in HTS Hubs where the number of controllers is proportional to the number of the beams and each controller needs redundancy. The problem can be partially solved by using N:M redundancy scheme, when one redundant module is used to back up a few main controllers. However, it is applicable only when all controllers have the same connection scheme, which is not always the case in HTS Hubs that work in two polarizations. Each group of controllers requires its own redundancy. There it seems impossible to eliminate a major complexity/cost associated with Hub redundancy. Or are there any more innovative solutions?

Smart redundancy

Good ideas can be found nearby – in every data center. Cluster computing systems are famous for theirhigh reliability with smart redundancy due to load balancing, i.e. dynamic redistribution of tasks between similar computers. If one computer in a cluster fails, its load is picked up by any other computer with spare processing resource. Is such an approach feasible in a satellite network?  It certainly is, provided the network in question is designed with modern software-controlled satellite routers!

UHP VSAT platform is one of the most recent technologies to arrive on the market, but thanks to its innovative nature it has already fundamentally changed many traditional approaches and solutions in the field of satellite communications. No wonder that we have the most disruptive views of redundancy. 

Without a doubt the Hub is the most expensive component of a VSAT network.  It consists of the hardware elements and software that makes the whole system work. In a modern high-technology system the ratio of software vs hardware costs can reach 90%/10%. The distinctive characteristic of the UHP software-defined routers is their high universality. Any router can play different roles depending on the software, for example can be a remote terminal or a controller, which the basic building block of the UHP Hub. Most importantly, the required functionality can be easily and automatically transferred from one UHP device to another. This unique feature of UHP routers is the basis of the innovative solution known as UHP Smart Redundancy.

Architecture of the Hub with Smart Redundancy is not different from a traditional Hub architecture, but all the controllers are universal and can assume any role. The network management system dynamically assigns specific roles to the universal controllers.

Software-defined functions or profiles, embedded in the Hub controllers, represent the major part of the total Hub cost. Transforming the controller profile into a virtual unit, which can be assigned dynamically to any universal controller, significantly reduces the system cost without compromising its reliability. Some extra hardware in the central stations still required, but that does not have a strong impact on the total cost of the Hub. Moreover, the amount of redundant equipment is limited because hardware controllers are universal as the above diagram illustrates. 

How does it work?

The controller functionality keys are initially installed not in the Hub equipment itself, but in the Network Management System (NMS). The universal controllers work under management of NMS that at launch selects the right number of available controllers and assigns to them functions required in order to achieve a specified network configuration.

The controllers that do not receive any current role are considered redundant. The NMS continuously analyzes status of the network.  If any problems arise, it re-assigns the role of a failed device to one of the redundant controllers. The re-assignment makes use of the same software key that was earlier used in the failed device. That way the software is always in use without any downtime.

The NMS will “fight” for survival of the VSAT network until the last Hub controller. Even if there are no more idle (standby) controllers available, Smart Redundancy does not stop working, but only changes its tactics. Priority levels can be preliminary assigned for each role in the network. For example, a forward channel controller should have the highest priority, as the complete network depends on it. The return channel controllers can have a lower priority because the network can continue to operate, even if some of the return channels are unavailable. In this case the network throughput in the return direction will decrease however the service is maintained. Some controllers in the Hub may have less important roles in the network, e.g. control terminals or SCPC-DAMA controllers. All these devices can be considered by the NMS as a pool of resources, and their role can be changed in case that a controller with a higher priority fails. Such flexibility and intelligent “self-healing” can only be afforded by the UHP Smart Redundancy and is not available with traditional redundancy approaches.

Geo-redundancy

Many HTS satellites use Ka-band which is sensitive to atmospheric conditions. As a rule, this problem is solved by Hub Site Diversity, that is adding another teleport in a separate location, which takes over when the main teleport cannot provide the requested service (for example, because of a rain fade). Geographic redundancy, or geo-redundancy, is an important requirement of SCADA and IoT networks where a single point of failure in the network is not acceptable and high service availability is a key requirement.

UHP Smart Redundancy is an optimal decision for Geo-redundancy. Both teleports have compatible equipment, managed by the same NMS. At the first level of redundancy the NMS tries to maintain normal operation of the main teleport by means of the existing local controllers. If the problem cannot be fixed at this level, for example, if the main teleport is subject to major propagation problems in the satellite link or is out of redundant controllers, the NMS decides to transfer all the services to the remote teleport.

The second level of redundancy switches all the controllers of the failed teleport to the backup teleport. It is worth noting that the UHP Smart Redundancy scheme uses the same unified functionality keys for both local and geographic redundancy that gives a fourfold economic effect in comparison with the traditional geo-redundancy approaches. In addition, the UHP redundancy approach has a simpler and hence much more reliable architecture.

The UHP Smart Redundancy scheme is simple and flexible. It also makes it possible to expand and evolve software embedded in the Hub independently from the Hub hardware. The number of universal controllers in the Hub can increase or decrease in the course of Hub operation. These controllers are based on the Universal Hardware Platform, same as used in the remote terminals. Thus, the controller hardware is low-cost and readily available. Activation of an additional network controller is done via simple installation of the required software key into the NMS. Any available standby universal controller can adopt this new role without compromising the active redundancy scheme.

Satellite networks have unsurpassed survivability and resilience in presence of natural or anthropogenic disasters, as many companies dealing with critical infrastructure have realized. Terrestrial networks are greatly impacted by such disasters, so adding a satellite component can greatly improve the overall network reliability and availability. UHP Smart Redundancy takes the network availability to a new previously unattainable level while only requiring a very modest investment from the network operator.