In contrast to traditional DVRs, cloud based DVRs enable providers to offer greater viewing options, more storage and flexible recording without the need to deploy hard-drives at the subscriber site. Ericsson’s cloud DVR solution does this through its software-based infrastructure, the Ericsson Video Storage and Processing Platform.

The platform’s infrastructure allows for seamless augmentation and replacement of legacy TV services with new cloud-based services. By virtualizing the storage and processing capabilities of as many Commercial-Off-The-Shelf (COTS) servers as needed, it prioritizes performance and allows operators to avoid many of the complexities and costs associated with launching new services.

Ensuring scalability

For all cloud DVR applications, scalability and capacity efficiency are critical, and these cannot be affordably achieved using common storage and streaming server architectures. Ericsson’s solution scales linearly to overcome this challenge, and can scale-out to support hundreds of petabytes of video storage as well as great numbers of concurrent recordings and playouts. The platform achieves this by integrating storage and processing resources. This leverages all resources to ensure optimum performance, and eliminates the need for independent server silos for encoding, streaming and so on.

For subscribers this scalability means more space for additional recordings without the need for a truck roll or hardware upgrades. It also addresses the problem of multiscreen viewing, which demands that multiple file formats are saved, dramatically increasing storage needs. Since storage capacity is dynamically allocated with this solution, consumers can upgrade their storage capacity simply without any upgrade of their device, and no longer experience limitations on the number of concurrent recordings that can be performed at once.

Cost-efficiency

The platform has been designed so that it can achieve massive scale whilst enabling cost-efficient deployment of private copy, shared copy, or hybrid systems. First, the platform is hardware agnostic, meaning that operators are free to choose their own generic x86 servers and still expect heterogenous support, thus reducing the need for expensive appliances or vendor lock-in.

Secondly, by integrating compute and storage into the same stack, packaging, transcoding, streaming and ad insertion are all integrated. In fact, the stack is optimized for media workflow, negating the need for ‘silos’ or dedicated tracks for different functions (one for streaming, one for encoding etc.) This dramatically reduces the server requirements and in turn OPEX.

Thirdly, the scale-out nature of the platform ensures that providers only expand as they need to. This eliminates the need to purchase in advance and then maintain large, under-used racks of servers in anticipation of growth that may or may not occur.

Performance:

The platform has been built to sustain very high recording and playout concurrency, to help compensate for the potential 30-40% recording concurrency at peak times, such as prime time, or during popular programs. The unified storage and computing platform is built using the converged cloud architecture resources principle, leveraging all the resources in the grid. In addition, superior load balancing ensures there are not bottlenecks.

Consumers are increasingly coming to expect a limitless ability to record programs, from any of their devices, and Cloud DVR can answer these needs. Ericsson’s platform design means that scalable technology has been carefully integrated with an architecture that maximizes cost-efficiency. This way the heavy burdens of CAPEX and OPEX, which are associated with traditional DVRs, are dramatically lifted. By eliminating expensive DVR-enabled set top boxes and decreasing the need for hardware replacement and upkeep, providers can expand their offerings and add new revenue streams while reducing costs.