Broadband over Power Lines (BPL)

Introduction
Broadband over Power Lines (or BPL) uses the existing power grid infrastructure to provide high-speed, broadband Internet access to homes and businesses. It is a new innovation based upon existing Power-Line Communications (PLC) technology.

Figure 1: PLC A History, based on information from [1, 15].

Advantages
Despite the proliferation of broadband access methods in many countries, some advantages could still be attained by rolling out BPL. Since it uses the existing infrastructure, BPL could mean that low-cost broadband could be made a reality in areas that cannot get DSL, cable or wireless broadband. Even homes in extremely remote areas could now potentially get broadband, without having to resort to the high latency satellite broadband.

Since BPL potentially provides every household with an alternative to the telephone network over the last mile with no extra cabling required, it brings much-needed competition to the often one-horse local loop ownership market. Therefore, the technology offers advantages even for urban dwellers. Another potential benefit of BPL is the possibility of its use for smart appliances [1]. The idea behind this is that you can control appliances with your PC. While these devices could potentially be connected with Ethernet to a DSL connection, BPL offers a much neater solution, since a single plug acquires both the devices electricity and its data. Some propose this as an aid for people with mobility problems.

Technology
Fundamental Concept
Since BPL uses the existing power grid infrastructure, power lines now have two purposes, both of which must coexist without interference. It the late 1980s, it had been observed that a fixed telephone lines local loop used only a fraction of its available frequency range. This led to the invention of DSL, which exploited the unused frequency range to provide broadband Internet access. Standard AC electricity is transmitted at a frequency of 50 Hz or 60 Hz. Researchers noted that this left almost the entire frequency range of the line free, which suggested that perhaps, like the local loop, the line could be used for additional purposes. Consequently, it was proposed to transmit data over the unused frequencies of the power lines, using methods similar to those used for DSL. This is the foundation idea upon which BPL technology is constructed. However, transmitting data over power lines for extended distances leads to degradation of the signals. This problem is particularly exacerbated by high-voltage power lines. Consequently, data is typically transmitted over fibre optic cables for the length of the high-voltage lines, before being converted to electrical signals and being added to the medium-voltage or low-voltage lines. This is rarely a problem, since the majority of electricity transmission companies, including ESB Networks, have wrapped fibre optic cables around their high-voltage lines. However, even on medium- and low-voltage lines, boosters are required at regular intervals (as with most other technologies) for reamplification to prevent signal loss. Once the signals have arrived at their destination, there are two options for transmitting the data into peoples homes: wired or wirelessly [2]. Most companies are primarily considering the wired option, whereby a special modem is plugged into a standard electrical socket. Other companies are designing their products based upon the current enthusiasm for wireless communications, placing base stations based upon IEEE 802.11b (or other wireless networking standards) on electrical poles, which may then be received using standard equipment.

A pioneer of BPL, Media Fusion amazed customers and ISPs with promises of speeds up to 2.5 Gbps [3]. However, to date, the company has failed to bring such promises to market [4]. Instead, speeds of approximately 13 Mbps are standard [5]. Such a system would quickly saturate but manufacturers are working on improving the technology to provide connections of greater speed, potentially making the technology viable.

Technical Details
Since extremely few BPL systems are commercially available, there has been no attempt to standardise the underlying protocols. Instead, trials have typically tested a multitude of technologies, in an attempt to find the most appropriate one for the harsh data transmission medium of a power line [6]. It should be noted that solutions designed for BPL are often based upon those used in mobile communications, since both technologies could potentially suffer from high error rates and, consequently, from low data rates. The two main choices of technology used to implement BPLs physical layer are CDMA (Code Division Multiple Access), which is used in some mobile telephone systems, and OFDM (Orthogonal Frequency Division Multiplexing), which is used in IEEE 802.11a. Both exhibit favourable characteristics, although performance studies indicate that CDMA performs substantially better than OFDM in terms of the data rate achieved [7]. However, if the line were very noisy, OFDM would perform much better. Consequently, OFDM is more fault-tolerant. It is, perhaps, for these reasons that OFDM has been more widely researched than CDMA. Further, the ESBs trials were conducted using OFDM [8]. Additionally, CDMA is not currently viable due to difficulties in creating enough CDMA chips. Work is already in progress to overcome these limits, however. Two conditions must be considered when designing the MAC (Medium Access Control) sublayer of BPLs data link layer: there is no limit to the distance between nodes and multiple nodes may transmit simultaneously. The first condition eliminates the CSMA/CD (Carrier Sense Multiple Access with Collision Detection) protocol used with Ethernet. However, the CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance) protocol used with IEEE 802.11 is suitable and is a widely researched solution. The Bluetooth protocol is also suitable. However, as could be expected, CSMA/CA performs better, as its design goals are a closer match to BPLs [9].

Figure 2: Illustration of the collision avoidance phase of the IEEE 802.11 (CSMA/CA) protocol [10].

Figure 3: Piconets, the fundamental idea behind the Bluetooth transmission protocol [10]. Higher-level layers are of lesser consequence, as no special technologies are usually required for the medium.

Problems & Solutions

The power lines would need repeaters to maintain signal integrity [1] and since the data signal cannot pass through transformers (in which case it would be lost), they must be bypassed. Routing data around transformers can be costly [11]. Since power supply networks vary from country to country, the cost of transformer bypassing can vary. To generalise, houses take in a low voltage (LV), so the medium voltage (MV) used for transmission must pass through a MV/LV transformer before it can enter a house. In the US, 1-10 houses are served by a MV/LV transformer, in Japan the figure can be up to thirty, while in Europe several hundred houses can be serviced by a single transformer [12]. This may account for the fact that BPL has been made commercially available in some European countries [13], while in the US utility companies are still engaging in trials. On the other hand, a report by the National Exchange Carrier Association [1] estimated that it would cost $10.9 billion to lay the wiring needed to provide rural areas in the US with (conventional) broadband. Just because BPL would be a cheaper alternative does not mean it is economically viable. The cost of transformer bypassing is not the sole economic headache for potential providers. Since powerlines were never intended to be used for piggybacking data [12], a number of problems arose when trying to do so. These include high attenuation at high frequencies and noise (internal and external) [14]. As has been mentioned earlier, this leads to the necessity for a lot of error correction/prevention in any protocols using power lines as a physical layer. One thing that cannot be resolved, however is a failing in the electrical properties of the powerlines themselves. They act as aerials because they are

not shielded [15]. This means that they can pick up noise and transmit it on as well as emit interference. Unfortunately, BPL operates at the same frequencies as short wave radio and low-band VHF. This can render various radio systems including those of governments unusable [16]. Amateur radio enthusiasts the world over seem to be united in their distaste for what BPL does to the airwaves [16, 17]. This interference has historically scuppered BPL trials. A good example of this is the Nor.Web trial that began in 1998 in Manchester[15]. Despite complaints about the interference and warnings from the Radiocommunications Agency [18], the company consistently rubbished criticism and insisted that the roll out would take place. By the end of 1999, the company had been closed down [15]. In Japan, the technology will not be adopted because of the interference problem [10]. Current trials seem to be suffering from the same problem. Power company Scottish Hydro Electric is currently offering BPL in three towns for 35.99 and 29.99 per month for 1 Mbps and 512 kbps connections respectively [19]. While the price is in the region of DSL, the interference problem has not gone away: BBC engineers have confirmed this [1]. Even in Germany where there are many companies offering commercially available BPL, the University of Duisburg-Essen has had similar findings when testing interference levels [15]. In the US, the FCC has approved guidelines for the implementation of BPL [20]. This means that there is a set limit for acceptable radio emissions from the technology. Many US trials have found BPL financially unviable. Others currently taking place cannot meet the FCC requirements restricting radio emissions [14]. In other countries, there are no guidelines for what is acceptable. Electricity companies have a steady core business. Why invest heavily in an unproven, unregulated technology that could be shut down as soon as a government realises that it is time to regulate or even ban the technology? The shareholders would not be very impressed. Another problem with BPL is security. Since it transmits on a shared medium, like cable broadband, this makes it easier to snoop the line. Even though European operators have to spend less on transformer bypasses as has been already explained, the fact that the LV signal can potentially go to several hundred homes is not very secure. The same line going into many homes means the same traffic going down that line. This provides an opportunity for hackers to acquire sensitive data. The only proposed solution to the radio interference BPL causes is one proposed by Corridor Systems [21]. They propose to use microwaves instead of the lower frequency bands to transmit the data, meaning that radio equipment should not be interfered with. Supposedly, this could lead to data rates of up to 216 Mbps. In the US, the National Association for Amateur Radio (or ARRL, which has been one of BPLs most vehement critics) has acknowledged that such a technology would not interfere with radio signals used by amateur radio enthusiasts. The electromagnetic spectrum is quite congested [22], however, and using the 2-20 GHz bands may just spawn more opponents to BPL. Radio astronomers, who make use of several protected frequency bands from 13 MHz all the way up to 275 GHz [24] may be BPLs next opponents. Given that the 1-10 GHz bands

are especially important in this field of study [23], and that Corridor Systems 2-20 GHz BPL has not yet undergone extensive trials (or even been implemented?), we can only speculate at this time.

Figure 4: Who uses the EM spectrum?, based on information from [22].

Figure 5: BPL acts as an aerial [17].

Conclusion
The next few years will decide whether BPL can compete in the broadband market. The quantity of research in the field has resulted in the solutions outlined above. These solutions are the foundation for making BPL viable. But even if it is technologically viable, will it be economically viable? BPL offers a method of broadband access for those living in isolated areas, who have no other viable means of broadband access. Therefore, it seems plausible that when BPL will become available in rural areas, it will be a moderate success. However, this success is unlikely to be long-term, since telecommunications companies are already contemplating rolling out FTTH (Fibre to the Home) connections to all of their customers sometime in the future. Therefore, it appears that BPL will be little more than a stopgap solution.

But will electricity companies support BPL en masse? Certainly, the possibility of the technology being shut down without notice, as outlined above, would be of concern to many of these companies. Moreover, the requirement of a network upgrade and the possibility that it will only be an interim solution would be a disincentive to many such companies. In fact, both Nortel Networks and Siemens have backed out of the technology that they previously claimed was the future of broadband, citing the costs of network upgrades [24]. Overall, BPL has a future albeit a limited one.