Skip to main content

New IEEE P1905 standard may unite Europe's fragmented home networks

The entry of the IEEE, the world's largest technology standards body, into the TV home networking arena highlights the coming of age of this field, while also suggesting that more work is needed to deliver interoperability. The IEEE is developing a standard called P1905, which will be an abstraction layer allowing devices and services to work over any of the four main existing home physical communication technologies — HomePlug AV over power lines, MoCA over coaxial cable, Ethernet over Cat 5 twisted pair cable and Wi-Fi for wireless. It will enable a pay TV service to work over a hybrid home network comprising a mixture of two or more physical layers, such as Wi-Fi and power line, through common management and control. Indeed P1905 will completely shield video management software from the underlying physical connectivity and allow the network to operate seamlessly over multiple physical technologies. It will avoid the need for a gateway or an application to know what physical network it is operating over, bringing the four technologies together within a single abstraction layer interface and be capable of using whichever option is available or performs best at a particular time.

The emergence of P1905 with the weight of the IEEE behind it does beg the question of why another home networking standard is required when there seem to be so many already making some claim of universal interoperability. Digital Living Network Alliance (DLNA) was supposed to facilitate interoperability within the home, but that is more of a coalition, incorporating standards developed by other bodies as well as adding some new components, in particular for automated device and service discovery.

More pertinently, G.Hn appears to have the same goals, and the existence of P1905 does seem a rebuke to that initiative. The IEEE argues that P1905 addresses some deficiencies in G.Hn by operating at a higher level. is a universal physical interface designed to integrate coax, power lines, Wi-Fi and Ethernet into a single physical network. The P1905 camp argues that there is no need for integration at the physical level and points out that G.Hn components would not be backwards compatible with existing MoCA, HomePlug or Wi-Fi ones. It also argues that G.Hn does not yield any performance improvements over MoCA and the other existing physical standards, and that it has come too late into the market to replace them with this umbrella interface.

There is also the more subtle argument that there is no need for a universal interface, since, in practice, a given home might have two separate networks operating in parallel but not connecting to each other at the physical level. There might be Wi-Fi in every room but served by a MoCA coaxial backbone for example. Therefore there is no need for coexistence at the physical level that G.Hn achieves, at least according to the P1905 argument.

P1905 on the other hand allows coexistence at the network level, which leads to the IEEE's argument that it will enable a greater level of resilience by combining two or more physical networks. However this would only really be an advantage in the case of a power line/Wi-Fi combination, since only these two are susceptible to coverage issues within homes. By combining the two, users could be virtually guaranteed of service anywhere within the home, since the probability of both simultaneously failing to deliver service to a given point would be small.

The G.Hn camp dismiss this last argument as spurious on the grounds that in practice few home owners would want to combine two flaky technologies in order to provide a higher assurance of quality. This is not to suggest that power line is flaky or that WiFi will not continue to improve, but in situations where both do have coverage holes within a home — because of irregularities in wiring, perhaps, in the case of Powerline — then you would be better off cabling with coax, or even Ethernet, to provide a robust backbone.

Ultimately the market will settle the respective fates of G.Hn and P1905. To some extent, the debate will revolve around the question of whether the convenience of having a single PHY in the shape of G.Hn will prevail over its difficulties, such as lack of backwards compatibility, and the fact that with current 40nm manufacturing technology for signal processing chips, the incremental cost of adding separate interfaces to consumer electronics equipment for MoCA, HomePlug and Wi-Fi is marginal. This is why P1905 is enjoying strong support from leading home networking silicon companies, such as Broadcom and Entropic, and that may prove decisive.

There is at least one point of consensus between the G.Hn and P1905 camps, which is that no one physical technology will predominate and that several will coexist, not just across the market as a whole but within many individual homes. Even MoCA has given up hopes of world domination, despite having won the home network physical layer battle in the U.S., where most homes have several coax sockets.