Readers of this blog already know that G.hn (also known as G.9960) is the new standard developed by ITU (the same group of experts that developed the highly succesful ADSL, VDSL and VDSL2 standards) for high-speed networking over existing home wiring (power lines, phone lines and coaxial cable). We have written about how fast G.hn is, about the various organizations that support G.hn, and about the major applications targeted by G.hn.
What we have not done before is providing a comprehensive overview of the most important advantages provided by the standard. So, that's the goal of this post.
#1: G.hn is faster than any existing wired home networking technology
G.hn will be fast. As we detailed in this post ("How fast can G.hn be?"), with its default setting G.hn can provide a line data rate over coaxial cable of up to 843.75 Mbps (with actual Ethernet throughput at 762.05 Mbps), and includes optimizations that can support line data rates up to 1.02 Gbit/s.
This does not mean that all G.hn products will operate at 1 Gbit/s. In fact, it's very likely that initial products will support lower data rates, and that future products will increase the maximum data rate, all without having to change the standard and without breaking interoperability.
Not only that: G.hn supports the concept of "profiles", which will allow silicon vendors to design devices with lower complexity and lower data rates for specific applications. We will see G.hn-compatible devices for applications like Energy management and Home automation, operating at low data rates, while maintaining compatibility with high-speed G.hn products.
#2: G.hn works over any type of home wire
Unlike other specifications that only support one type of wire (power lines only, or coaxial cable only), G.hn specifies a unified Physical Layer and Data Link Layer that can operate over multiple wire types (power lines, phone lines and coaxial cables).
This blog post ("Why do we need a unified standard at all?") explains the benefits of having a single PHY/MAC standard for multiple wires, and why Service providers have pushed for G.hn to support it.
Does that mean that G.hn will not be optimized for any specific wire? No. In fact, G.hn includes parameters that are specifically optimized for each medium. For example, key elements of the PHY/DLL, such as OFDM sub-carrier spacing, the Forward Error Correction (FEC) or the ARQ (retransmission) mechanisms.
G.hn will outperform any of the existing wired home networking technologies over any medium.
#3: G.hn is supported by multiple silicon vendors
Unlike with existing wired home networking technologies, which traditionally only have had a single silicon vendor providing actual products (severely limiting customer choice), multiple silicon and IP vendors (including DS2) have already announced support for the G.hn standard in their future products. And we are not talking about start-ups here. These are serious vendors who have shipped several millions of chips into the home networking and DSL markets.
Having multiple silicon vendors offering interoperable G.hn products will create healthy competition in the marketplace, will accelerate technical innovation and will ensure that customers get the best products at the best possible price.
#4: G.hn is supported by multiple Industry Groups
HomeGrid Forum is the organization set up specifically to promote adoption of G.hn and to ensure interoperability and compliance with the standard.
But HomeGrid Forum is not the only organization supporting G.hn: On February 2009, three home networking organizations that promoted previously incompatible technologies (CEPCA, HomePNA and UPA), announced that they had agreed to work with Homegrid Forum to promote G.hn as the single next-generation standard for wired home networking, and to work to ensure coexistence with existing products in the market.
#5: Most products based on G.hn will provide compatibility options with existing home networking technologies
G.hn, by itself, is not directly interoperable with existing wired home networking technologies. There is a reason for this: there are at least 5 completely different legacy specification. As this article says ("On the issue of G.hn's FEC"):
If G.hn tried to be compatible with all of the existing wired technologies that have shipped millions of devices into the market, then it would have 5 different modulation schemes, 5 FEC schemes, 5 security schemes, 5 MACs, 5 of everything. That would be the best way to make it the most complex standard ever designed. But G.hn is about simplicity (one PHY and one MAC that works anywhere) so the group decided early on that they didn't want to follow that path.
So, what about the installed base of existing products? Do we have to replace them? Fortunately not. Those G.hn vendors with an installed base of legacy home networking technologies have announced plans to develop "dual mode" chips that will be compatible with G.hn and with legacy specification. DS2 was the first vendor to announce this, which means that products based on the UPA specification will interoperate with future G.hn products from DS2.
#6: G.hn provides state-of-the-art security
G.hn uses AES-128 as the encryption algorithm, and ITU Recommendation X.1035 as the protocol for authentication and key exchange. G.hn security is very strong (much stronger than that provided by many existing systems based on DES and 3DES), and provides an additional advantage to system designers over many other home networking technologies: G.hn products don't need to support a plethora of legacy encryption mechanisms (unlike IEEE 802.11 products, which usually need to support multiple security schemes, such as WEP, TKIP and CCMP).
Having fewer options also means better security, as there are fewer chances to introduce bugs in G.hn implementations.
#7: G.hn will have longer range than most existing home networking technologies
G.hn includes a nice feature specifically designed to extend the range of the network: relaying. Although DS2 products based on the UPA specification have supported this feature for many years, most existing wired technologies do not support it.
Automatic relaying is a key technology for wide area networks such as Broadband over Powerline networks deployed over utility power lines.
Using this feature, a G.hn "source node" can use an intermediate "relay node" to send data to another "destination node", even if the source and destination nodes are not within direct reach from each other. This feature improves network reach and will allow G.hn to be used in large installations.
#8: G.hn will reduce energy consumption
G.hn will include mechanisms that will allow devices to go into "sleep state" in order to reduce energy consumption and to quickly get back to "active state" as soon as a device needs to send data.
Advanced support for "sleep states" are required to support the latest European Code of Conduct on Energy Consumption of Broadband Equipment.
#9: G.hn will provide reliable communication over noisy home wires
G.hn includes multiple mechanisms to improve reliability over any kind of wire. Of the three wires supported by G.hn, power lines are probably the harshest ones, so the G.hn group has spent a significant amount of time optimizing performance for that case.
How does G.hn handle reliability over noisy power lines?:
- Forward Error Correction (FEC): G.hn uses a state-of-the-art LDPC code to protect data transmission. Before transmission, LDPC codes add redundancy to the data, which is then used by the receiver to recover the contents of the data even if some of the bits have been corrupted by noise. G.hn selected LDPC (over other options) because it provided the lowest Block Error Rate over the expected range of operation, and because LDPC decoders are easier to implement at high data rates due to their inherent parallelism (which is one of the reasons why LDPC is also used in other high data-rate standards like 10GBase-T Ethernet)
- Selective ARQ: G.hn implements an ARQ (Automatic Repeat Request) mechanism that re-transmits data frames that have been affected by too many errors.
- Synchronization with the AC cycle: Noise in powerline is frequently synchronous with the AC cycle. This means that if the AC cycle has a frequency of 60 Hz, noises generated by electric appliances also have a frequency of 60 Hz, or some times twice that (120 Hz). If a device detects a strong noise spike, it's very likely that the same noise spike will show up 1/60 = 16.6 ms later. A G.hn device can take advantage of this and schedule its transmission to avoid this "predictable" noise.
#10: G.hn will provide predictable service to QoS-sensitive applications such as IPTV
Many existing wired (and wireless) networking systems use Medium Access Control mechanisms based on variations of CSMA/CA. One of the advantages of CSMA/CA is its simplicity, but this comes at a cost: because CSMA/CA is collision-based, performance of CSMA/CA-based systems is very dependent on network load and QoS in general cannot be guaranteed. Systems like CSMA/CARP improve this by introducing priority-based access, but the problem still persists when multiple system with the same priority want to use the channel at the same time.
The G.hn MAC is based on a master/slave TDMA architecture, in which a central device ("the domain master") allocates channel access to other "slave" nodes in a predictable manner. Slave nodes can request specific allocations of bandwidth to the domain master, which can implement them by assigning exclusive "contention-free" time slots to each slave.
With this mechanism, G.hn can provide guaranteed bandwidth and latency to applications that have strict QoS requirements, such as IPTV, VoIP or on-line gaming.
| Request more information on DS2's G.hn products for next generation home networking over power lines, phone lines and coaxial cables |
Related blog posts on G.hn:
- In-Stat and HomeGrid Forum publish webinar on next generation G.hn networks
- How fast can G.hn be?
- HomeGrid Forum publishes white paper on G.hn compatibility with existing home networking technologies
- DS2: Driving Industry Standards
- G.hn keeps making progress
HomeGrid Forum resources on G.hn (free registration required):
How to get technical specification of G.hn 9960, in case of not having an ITU TIES account?
Posted by: John Chang | October 05, 2009 at 03:16 AM
@ John Chang,
Hi John. Once G.9960 is approved, it will be published here: http://www.itu.int/rec/T-REC-G/e
In the mean time, the only way to get a copy of the draft is having a TIES account at ITU.
Posted by: Chano Gomez | October 05, 2009 at 03:52 AM