Spread Spectrum Radio.
© Mercury Communications Ltd - August 1992
The spread spectrum approach to wireless data transmission originally derived from the need by the military for secure and reliable radio transmission on the battlefield. As such, it has been an established transmission methodology used for both for radar and radio applications for many years.
Only in recent times has spread spectrum transmission been considered for use in commercial environments. The three principal application areas being: (1) intra-office LAN interconnect i.e. wireless-LAN, (2) cellular radio and personal communications systems and, (3) local access applications. As wireless LAN technology is to be discussed in a future issue of Technology Watch, a good place to start is to have a brief look at digital cordless telephone and cellular radio standards.
Digital Cordless Telephony
Within Europe there are two standards for digital cordless telephony: telepoint CT2 and digital cordless telecommunications, DECT.
CT2 and DECT
In the early 1980s the European analogue cordless handset market was threatened by a potential flood of imported units. Responding to this, Europe developed three standards of its own. Outside of the UK and France, the Conférence des Administrations Européenes des Postes et Telecommunications, CEPT, CT1 standard was adopted. This offered good speech quality but was expensive. In consequence, both the UK and France developed their own, incompatible, standards. Although CT1 and all its variants were analogue based, it was soon realised that a common European digital standard would be beneficial. In the UK, the CT2 standard was initially developed by a number of companies under the sponsorship of the DTI. It has subsequently been revised with Common Air Interface and adopted by some CEPT and 3rd world countries.
CT2 uses 40 separate frequencies (carriers) to provide 40 duplex channels i.e. a frequency division multiple access, FDMA, approach. Each pair of send and receive channels are multiplexed onto a carrier in a time division multiple access, TDMA mode. The result is a combination of TDMA and FDMA techniques as shown in Figure 1. One major benefit of CT2 over DECT technology is that it is available today.
Figure 1 - CT2 TDMA Frame Structure
As the movement for the establishment of a common European digital standard grew, a debate started encompassing, technical, commercial and political aspects between the two rival technical approaches. One was based on time division multiple access, TDMA, and the other based on frequency division multiple access, FDMA. Throughout 1987 discussion went on, and in early 1988 a compromise was reached in the form of the Digital European Cordless Telecommunications , DECT, standard operating at around 1800MHz.
DECT arose from the failure to reconcile the differences between the FDMA and TDMA protagonists and so compromise resulted in combining both approaches. DECT provides 120 duplex channels using 10 separate frequencies and multiplexing 12 send and receive channels on each. Although DECT technology provides more comprehensive facilities than CT2, it will not become available until late 1993 or early 1994.
Digital Cellular Radio
Digital cellular radio has also been moving toward the adoption of a common digital standard known as Groupe Spécial Mobile, or GSM. GSM will enable European-wide compatibility for the first time. GSM uses the 900MHz band, and as with DECT, uses TDMA but with eight time slots per channel. Hence one channel can support eight full rate or sixteen half rate channels. Channel separation is 200kHz with mobile transmit channels in the range 890 to 915MHz and mobile receive channels in the range 935 to 960MHz.
GSM has been supplemented in the UK by an additional cellular service called the Digital Cellular Service, DCS 1800, based on GSM technology. As a modified form of GSM, it operates in the 1800MHz band with smaller cell sizes and lower power. Phase 1 of the standard, finalised in January 1991, stipulates a number of differences to its forebear at the base station and handset level. As the standard is mainly aimed at personal communications network, PCN, applications, it has been optimised for higher-density traffic than would be seen in smaller PCN cells.
To reiterate, FDMA separates channels by putting them on different frequencies, while TDMA separates them by allocating different time-slots on the same frequency to a sender and receiver. In practice, GSM, DCS 1800 and DECT all use a combination of these techniques to multiplex users.
In the USA, similar discussions have taken place between these alternate technical approaches, but with the addition of one other, code division multiple access, CDMA. CDMA is a spread-spectrum approach to user multiplexing. Although CDMA has found a definite niche in a commercial guise in the wireless-LAN market in the US, has temporarily lost the battle to TDMA/FDMA in cellular and PCN arenas. So far in Europe, there has been very little activity with regard to spread spectrum transmission outside of military use.
Spread Spectrum Approach
To understand how spread spectrum transmission operates it needs to be contrasted to conventional, narrow-band, point-to-point or point-to-many radio transmission.
The Narrow Band Approach
Conventional narrow band radio transmission uses transmitters and receivers tuned to the same, single, frequency. In a cellular network, it is possible for a receiver to scan a number of channels looking for activity. Once an active channel is detected, the receiver ceases scanning and data exchange takes place on the selected pair of transmit and receive channel frequencies. While in communication, the transmit and receive frequencies do not change. This approach has it strengths, but it also has its weaknesses. For example, adjacent cells in a cellular network need to be allocated different frequency groups to avoid interference and transmitter output power needs to be limited for the same reason. Also, without encryption, it is easy for anyone to eavesdrop on the link. CDMA spread-spectrum transmission can overcome these weaknesses.
Although considered a new technology on the European commercial scene, spread-spectrum transmission has had a long history in the military communications market. Its main benefit being that spread spectrum communication systems permit tactical battlefield communications that are not detectable by enemy jamming and monitoring systems.
With the decrease in the amount of spending on defence in the western world, many military equipment manufacturers have been looking for commercial applications for technology developed for the military. With the increased need for commercial communications in the 1990s, comes a need for efficient use of the frequency spectrum. Spread spectrum communications results in efficient spectrum use. In the USA its use has already been widespread, particularly in the wireless-LAN product area based on frequency allocations given by the Federal Communications Commission , FCC, dedicated to spread spectrum activity. Allocated frequencies are 902 - 928MHz, 2400 - 2483MHz, and 5725 - 5850MHz. Many new small companies were set up to take advantage of these allocations and there are now literally thousands of users. Almost all of this activity is with wireless-LAN products.
Little activity with spread spectrum has taken place in Europe due to a lack of frequencies allocated to this use. The greatest concern being that of possible interference to existing users.
Characteristics of Spread Spectrum.
A spread-spectrum radio transmission system, rather than using two dedicated frequencies (or a pair of channels) for duplex operation, spreads the power to be transmitted over a wide band of frequencies. This is similar to the carrier-hopping techniques employed in frequency agile RADAR. When this is accomplished, the transmitted spread spectrum power received by a conventional user having a narrow bandwidth, is only a small fraction of the actual transmitted power. Thus an ordinary radio receiver would only detect an increase in the background noise when placed near a spread-spectrum transmitter. For example, if a spread spectrum signal having a power of 1 watt is spread over a bandwidth of 100MHz and an existing band user employs a narrow band communication system having a bandwidth of only 1MHz, then the effective interfering power in the narrow band system is reduced by a factor of 100 to 10mW. This allows spread spectrum users to co-exist with conventional users, and for many spread spectrum users to co-exist on the same frequency band.
For any spread spectrum system to work, it is necessary for the receiver to acquire and continually track the phase (or frequency) of the transmitter i.e. the receiver needs to synchronise itself with the transmitter. Initiation of a spread spectrum link requires that the receiver undertakes a search until a high correlation is achieved between the incoming signal and the locally generated internal spreading code or sequence. If synchronisation cannot be maintained, the desired signal cannot be detected.
There are several ways data can be encoded in spread-spectrum transceivers, the Direct-Sequence, DS, approach is most common. With this technique, the carrier frequency is mixed (or more correctly modulated by) a pseudo-random binary sequence, or PRBS. This is a unique digital data sequence that repeats itself, say, every 1023 bits. If this pseudo-random code is added to the base-band data and mixed with the carrier, the transmission is effectively spread over 1023 differentfrequencies. Although close to being totally random in nature, when a pseudo-random generator is reset to zero it always cycles through a known sequence of ones and zeros. If both the transmitter and receiver use the same code, and synchronisation is maintained, the receiver will always know on what frequency the transmitter is operating. It is possible to generate an infinite number of pseudo-random sequences, so it is possible for many links to use the same band as long as they use different pseudo random sequences.
Code Division Multiple Access
Code division multiple access, CDMA, is a spread spectrum system in which two or more spread spectrum signals communicate simultaneously, each operating over the same frequency band. In a CDMA system, each user is given a unique sequence (pseudo-random code). This sequence identifies the user. For example, if user-A has sequence-a, and user-B has sequence-b, a receiver wanting to listen to user-A would use sequence-a to decode the wanted intelligence. It would then receive all the energy being transmitted by user-A and disregard the power transmitted by user-B. Figure 2, shows CDMA in use in a cellular system.
Figure 2 - Spreading at Transmitter & De-spreading at Receivers
Some Benefits of CDMA
Although CDMA was rejected as the standard for the next generation of mobile communications in the USA, it still has many advantages over TDMA.
CDMA is less prone to deep multipath fading caused by transmissions arriving at the receiver that have followed different paths. i.e. one signal direct, and another reflecting off a large object. In fact, one approach in common use with CDMA systems, a Rake receiver, takes advantage of multipath, normally a major source of interference and signal degradation in other systems. In a Rake receiver, each receiver coherently combines the three strongest multipath signals to provide an enhanced signal with better voice quality.
It is fair to say that this short note does not attempt to fully address the pros and cons of CDMA versus TDMA, and in fact a balanced discussion would need to discuss the benefits of TDMA over CDMA.
Millicom Holdings (UK) Proposal
In June 1990 Millicom Holdings (UK) Ltd made a proposal to the Radiocommunications Agency to operate a broad band CDMA radio network in the 3700 - 4200MHz fixed and fixed satellite band. The proposal pre-supposes being able to co-exist with current users of the band on a band-sharing basis for a fixed public telecommunications service providing a local-loop wireless access service.
The proposed network uses broad band spread spectrum techniques to provide multiple access capability via CDMA. The service is aimed at private subscribers and small businesses and will use small, highly directional, roof-mounted helical or quad-loop antennas with about ten to fifteen users per node. The base station will use an omni-directional aerial with screening to give high directionality in the vertical plane. Millicom claim that CDMA techniques will minimise interference to existing users who are fixed-site broad band links and satellite down links.
Base station power is predicted to be 10 watts maximum, with the worse case customer power being 1 watt. Up to 15 CDMA channels will be used. Millicom claim that this model is backed up by extensive experimental PCN work in the USA. Further, the use of highly directional aerials together with the possibility of using specific notch filters in the base station will minimise interference to existing users. If needs be, they state they would be willing to establish transmitter exclusion zones in areas where filtering proves to be inadequate.
Millicom, is an ex RACAL group company, which specialises in military tactical mobile communications and would therefore be well experienced in the design, manufacture, and deployment of spread spectrum systems. Adapting such systems to a fixed commercial application as proposed here should be well within their capabilities. The principle activity would be to simplify and cost-reduce designs to bring them into a price bracket that could be afforded by targeted users. This task would be eased by the availability of an application specific chip-set. Several of these are now available, based on PCN experimental work in the USA, the most well known of which is manufactured by the Californian RF house Qualcomm Inc. Certainly, the complexity of a spread spectrum handset should be no more intricate or expensive than seen in a PCN handset.
With regard to the services Millicom could potentially offer, Computergram International in Jan. 1992 offered: "..without the benefits of a UK-wide network of waterways to host its network, Millicom intends to build a radio-based system offering both narrow-band and broadband services over local and fixed links. Despite the differences in medium, both companies (Sprint & Millicom) seem to have spotted a similar niche, concentrating in their announcements on the high bandwidth market, with video and the like being touted as possible applications. No doubt there is a market requirement here, but the choice of niche can also be seen as a pragmatic response to the companies' lack of a local network.
Spread spectrum is still a relatively new technology in a commercial environment with little practical experience behind it yet in either a fixed or mobile cellular radio application. In effect, many of its benefits are theoretical with few real field trials in urban environments to allow evaluation and to ascertain how it would co-exist with other services.
All this aside, it has found a definite niche in wireless-LAN applications and its potential strengths for use as an access technology will find wider use in coming years as experience allays some of the practical concerns. If not successful as a technology for current generation of cellular mobile and PCN systems, CDMA will certainly feature in next generation systems.