Spread ALOHA Wireless Multiple Access: The Low-Cost Way for
Ubiquitous, Tetherless Access to the Information
Statement of the Problem
The ability to access the national information infrastructure (NII) on a tetherless, broadband basis is often discussed. But today, the capability to interchange information with and through the network on a tetherless basis is limited and expensive. The promises of new tetherless access approaches for data, such as the current personal communication system (PCS) implementation approaches, have been only thatpromises. But a new wireless multiple-access approach will change this situation in a revolutionary way. This new wireless multiple-access approach is called Spread ALOHA and is being developed under U.S. Small Business Innovative Research grants from the National Science Foundation and from the Advanced Research Projects Agency of the Department of Defense.
Envision a robust nationwide packet radio data network in place with millions of users having tetherless, broadband user communications devices allowing easy, low-cost automatic interface into the public and private networks of the national information infrastructureanywhere, anytime. A user with a portable PC, a personal digital assistant (PDA), or another device containing a Spread ALOHA PCMCIA card or embedded chip would have instant access to a network of choice.
This paper describes the market need for a wireless multiple-access approach that offers robust, wireless multiple access to the NII at an affordable price. A strategic plan for implementation of Spread ALOHA architecture having an increase of two orders of magnitude in capability over existing wireless data multiple access approaches is discussed.
The Market: What People Want and Need
As the NII evolves, users will increasingly want and demand ready access to acquire information, send information, or communicate with each other. Some of this information may be manually or automatically generated or requested by a user or by an application or stored in a database for manual or automatic request.
NOTE: Spread ALOHA is a trademark of ALOHA Networks Inc.
Access to the NII will either be by wire (telephone, cable TV, etc.) or wireless. The wireless case for millions of users is the focus of this paper.
Wired data access to the NII and predecessors with capabilities adequate for certain applications and with distinct evolutionary paths or alternatives exists today. But wired data access has a fundamental limitation. It is not always where the users are, and it does not travel easily with the users. One has to "plug into another jack" whenever one moves. The wire is a tether. Thus, wireless data access can prove a boon to those users who want and need information access wherever they are. The ability to operate on a tetherless basis generates new power to use information for almost everyone. (Those who have experienced a good, tetherless computing or data access situation can testify to this.)
But wireless data access capability to interchange information with and through the network on a tetherless basis is limited and expensive today. To date, new wideband tetherless access approaches for data appear to be "vaporware"many have tried and none have succeeded. There are fundamental technology limitations in wireless multiple access that have led to economic limitations of these dreams.
The Market: Quantization
Characterizing a market in which one introduces a product of at least two orders of magnitude more capability than currently exists is difficult at best. The following discussion forms a probable baseline that would be the lower bound for the market addressed by Spread ALOHA architecture.
The addressed market for nomadic, tetherless computer networking employing Spread ALOHA architecture consists of a large fraction of the users of portable computing devices, such as portable PCs and PDAs. The reality is, however, that with the introduction of a breakthrough technology such as Spread ALOHA, a new emphasis on new devices and applications comes into play, which tends to stimulate and transform the market. ALOHA Networks expects that the PC/PDA market will be only the base for the nomadic, tetherless computing network market. New applications using a PCMCIA card (beyond PC and PDA applications) and embedded Spread ALOHA wireless technology will develop as well.
Even recent forecasts for growth of the mobile data market can now be revised upward:
The mobile data market in the United States will increase from 500,000users in 1994 to 9.9 millionusers in 2000with a compound annual growth of 64 percent, according toa new report from BISStrategic Decisions.… The "mobile professional"sectorprofessional employees who spend 20percent or more of their time away from their desksrepresent apotential user population of 31.9million by 2000.…1
The rebirth of the PDA market will hinge on re-positioning the devices ascommunications-oriented PCcompanions.… BIS predicts protable PC users will increasinglycommunicate using mobile dataservices. We forecast that 2.8 percent of portable PCs that arewirelessly-enabled (300,000 units in1994) will grow to 16 percent of the installed base, or 2.6 million units, in1998. The increased usage ofmobile data will be a direct result of improvements in availability,functionality, and pricing for services.We also expect that the mobile data market will grow at a compound rate of 80percent through 1998.Although at most 12 percent of PDAs are currently wirelessly enabled, thatpercentage will grow to 75percent by 1998 [1.7 million units from authors' graph]. The current lowpercentage rate reflects thedearth of low-cost mobile data alternatives. Once users have more costeffective options to choose from,the number of wirelessly-enabled PDAs will climb.2
From the above forecast made without knowledge of the breakthrough Spread ALOHA technology, it can be assumed that the market for user units (as opposed to infrastructure) will likely be about 4.3 million units in 1998 and will approach 10 million units by 2000.
Existing and Near-Term Marketplace Alternatives
There are several alternative wireless data service approaches available or nearly available today. Broadly speaking, the services can be broken into three general categories: one-way, packet-based services (primarily paging); two-way, packet-based services; and circuit-based services. Packet-based services chop the information that is to be sent into data packets, attach addresses identifying the transmitter and the recipient if needed, and send the packet over a channel that is shared by multiple users. Circuit-based services allocate a specific transmission channel to the end user when a transmission is requested, and that user holds that circuit until the transmission is completed.
The most common form of circuit-transmitted data utilizes cellular modems to provide mobile data communication over the existing analog cellular infrastructure. By incorporating sophisticated error correction protocols, these modems attempt to compensate for the relatively low line quality and allow portable computer or fax machines to communicate on a wireless basis. This approach benefits from the existing broad coverage of analog cellular and, except for the continuing problem of broken connections, is effective for large file transfers. But the quality of service has been generally low, and the price to the user has been high because of long set-up times, low data rates, and high cellular airtime pricing.
Wide area applications such as the ARDIS and RAM Mobile Data services are struggling. (According to a New York Times article,3 ARDIS reportedly has 38,000 users in 10,700 towns and cities and RAM Mobile Data has 17,000 users in 7,000 towns and cities). These services, which are employed where customer data must be captured regionally or nationally, do not live up to their initial promise because they are narrowband and therefore fundamentally expensive and limited in data rate.
Cellular digital packet data (CDPD) services by AT&T McCaw, GTE, and others, which use the analog cellular infrastructure in the United States, are moving toward implementation in individual cellular carriers' regions. CDPD transmits packet data in the limited voice analog cellular infrastructure. Though it employs a packet structure good for many data applications, there are two limitations: (1) access to the infrastructure requires much of the cumbersome multiple access required for voice uses and (2) data rates are limited to what fits into the analog voice channels of that infrastructure, which are less than ideal.
Metricom, Geotek, and Nextel frequency hop/time division and time division, multiple-accessbased services and new two-way paging services, such as MTEL's Destineer, are also being deployed. Metricom offers a narrowband approach of uncertain throughput today, and its peer-to-peer architecture limits the data volumes its network can handlenot a desirable trait for ubiquity. None of these systems can provide tetherless data access that is wideband, robust, user unlimited, and inexpensive.
Though broadcast transmission (the transmission of significant amounts of data from one to many) is a well-understood problem, the transmission of significant amounts of data from many to one (i.e., the network node) can be quite difficult. The new wideband Spread ALOHA wireless multiple-access technology, contained in forthcoming chips and boards for manufacturers of nomadic, tetherless computing products, allows robust, "bursty," tetherless access to the NII at an extremely low cost. The implementation of Spread ALOHA is simpler and far less expensive than any of the various PCS or wide area data network approaches defined to date.
Spread ALOHA is an advanced wireless multiple access technology that can provide the capabilities required for digital networks with large numbers of remote terminals. Spread ALOHA combines the proven simplicity and operational flexibility of conventional narrowband ALOHA wireless multiple access with the high bandwidth and high throughput of spread spectrum. (Conventional ALOHA multiple access was developed at the University of Hawaii and is employed in RAM Mobile Data's network, in the Inmarsat maritime satellite system, and in many VSAT networks, and it is the underlying basis for Ethernet.)
Spread ALOHA compares favorably to time division multiple-access (TDMA) and frequency division multiple-access (FDMA) approaches because its capacity is data limited, not channel limited as are TDMA and
FDMA. That is, capacity in the Spread ALOHA architecture is limited only bythe amount of data transmitted inthe network rather than by the number of users who access the network. Table 1compares Spread ALOHA to theother data limited approaches, conventional ALOHA and Spread Spectrum codedivision multiple access(CDMA). It illustrates why Spread ALOHA is the most efficient multiple-accesstechnique for large numbers of users.
The number of users in either conventional ALOHA or Spread ALOHA is limited only by the total data rate of the channel. Conventional ALOHA is low bandwidth and thus has a low data rate because of the practical requirements of maintaining a constant pulse energy as the data rate increases. In CDMA the practical limit on the number of users is cell constrained by the requirement to implement a separate receiver at the hub station for each active user. In the IS-95 CDMA standard for cellular voice, the maximum number of users per cell is less than 40.
Spread ALOHA can be viewed as a version of CDMA that uses a single, common code for all remote transmitters in the multiple-access channel. In a Spread ALOHA channel, different users are separated by a random timing mechanism as in a conventional ALOHA channel rather than by different codes. Since only a single code is used in a Spread ALOHA channel, only a single receiver is required in the Spread ALOHA hub station, rather than a separate receiver for each remote terminal as is required in CDMA. In addition, because of the elimination of multiple codes, many of the most complicated features required in a CDMA receiver can be removed. The elimination of unnecessary system complexity makes possible a degree of system flexibility, which can be important in today's rapidly evolving wireless markets.
For example, a Spread ALOHA hub station need only be capable of synchronizing to received signals, all of which use the same code, a much simpler problem than that faced by a CDMA hub, where the codes received are all different. In a Spread ALOHA hub station, packet interference can be eliminated by a cancellation process made practical by the fact that the interference terms generated by all packets are identical. And in a Spread ALOHA channel it is possible to select a spreading code that has only half as much interference as codes used in a CDMA channel.
Spread ALOHA can provide a throughput 100 times greater than any conventional ALOHA network now in operation and is much easier to implement than current wideband multiple-access techniques, such as CDMA. No other technology can provide robust data networking with large numbers of users. Spread ALOHA combines the proven simplicity and operational flexibility of a conventional ALOHA multiple-access channel with the high bandwidth and high throughput of a spread spectrum channel.
Analysis and Forecast
Spread ALOHA will make it possible to build a nationwide broadband packet-radio data network allowing easy, low-cost, automatic interface into the public and private networks of the national information infrastructure. A user with a portable PC, a PDA, or another device containing a PCMCIA card or embedded chip would have instant access to the network of choice. Smaller campus, metropolitan, or regional networks could be addressed initially as a way of beginning what could ultimately be a national network.
Spread ALOHA technology also holds great promise for data/voice PCS applications. Because of its lower cost, Spread ALOHA offers a potential for ubiquity that does not exist with other approaches for supporting data applications within PCS. However, the PCS market cannot be easily approached without an
established standard for Spread ALOHA PCS applications for voice and data. The establishment of a standard is expected to take some time, but ALOHA Networks, in cooperation with others, plans to begin these efforts in 1995, looking toward the second phase of PCS standards and equipment in 1998–99. At the bottom line, however, if a nationwide data network can be implemented before the acceptance of a standard in this area, Spread ALOHA will be positioned as a de facto standard. ALOHA Networks believes that its Spread ALOHA technology will ultimately form the basis for the most viable PCS data standard.
The economics of using the Spread ALOHA technology are compelling. Spread ALOHA technology allows for a simple implementation and has no user population limits. This simple implementation can lead to low-cost user units. In addition, the cost of implementing the hub or microcell facilities is lower for a large number of users than any alternative.
ARDIS, RAM Mobile Data, and GTE's CDPD services appear to price their services at about $0.50–$1.00 per kilobyte.4 This is essentially a user cost of $0.50 per second of use! AT&T McCaw Cellular has announced CDPD service prices ranging from $0.08 to $0.16 per kilobyte.5 The typical average monthly bill for these services has been estimated to range from $25 to $200.
ALOHA Networks estimates that Spread ALOHA multiple access can substantially increase network capacity as well as individual ''burst" transmission rates without significant added cost over other alternatives. The existing and planned data networks tend to be constrained to an operating rate of about 20 to 50 kilobits per second. With large network volumes, price per kilobyte could be reduced by one to two orders of magnitude. As a corollary, Spread ALOHA could allow a pricing of $0.50 to $1.00 per 100 kilobytes if one assumes demand is stimulated by such a substantial price decrease.
ALOHA Networks anticipates that by 2000 users of a high proportion of notebook PCs, PDAs, and other embedded microprocessors in portable platforms will expect to be able to communicate with remote points, public networks such as the Internet, private networks, or with the user's office. In fact, with a low-cost tetherless approach, the ALOHA Networks' Spread ALOHA technology could significantly stimulate the market for notebook PCs, PDAs, and other devices not yet conceived! With the anticipated growth of nomadic computing, this would translate into 10 million to 100 million user units (either PCMCIA cards or embedded) in 2000.
The nomadic, tetherless computing network is envisioned as a nationwide system, accessible from almost anywhere in the United States. This network could either be integrated with other networks or interfaced to other networks at various nodes. The hub or microcell stations would be spaced according to propagation characteristics in every area, similar to PCS or perhaps ARDIS or RAM Mobile Data. (In fact, the existing infrastructure of these networks could be employed for this.)
Spread ALOHA can operate in almost any frequency band. However, the system envisioned here would operate in a given, yet to be determined frequency band. The possibilities are (1) allocation of a new frequency band for this service, (2) use of the existing ISM bands, (3) use of the existing ESMR bands, or (4) use of the existing and future PCS bands. Since the radio frequency transmission is spread spectrum, the selected band must, of course, be suitable for such uses. The frequency band approach, which will allow the fastest implementation but yet allow the expected ubiquitous growth, should be explored.
The network infrastructure for a nationwide nomadic, tetherless computing network would entail microcell sites with the Spread ALOHA hubs. These hub sites would be trunked together using existing telecommunications company and interexchange carrier facilities. At selected hub sites, the network would be interfaced with the Internet and other Internet-like networks.
To establish ubiquitous coverage, from 2,500 to 12,000 hub sites will have to be established, depending upon the selected frequency band and coverage patterns. A Spread ALOHA installed infrastructure of this size is estimated to cost between $100 million and $500 million. The exact cost depends on what the coverage pattern is, whether the data network overlays another network, and whether a voice network is implemented simultaneously. Such a network can be implemented on a phased basis, covering the highest user population areas first.
Remote User Terminal Communications Devices
ALOHA Networks envisions remote user terminal communications devices that are small and inexpensive. Assuming large quantities of devices, ALOHA Networks estimates the cost of the Spread ALOHA chip set or chip for remote user communications cards in a microcellular system to be substantially below $100 in the 1998 time frame, with the normal "Moore's Law" cost reductions beyond 1998. User software in the terminal device would employ "standard" user software such as the General Magic or other user operating software products. Ultimately, ALOHA Networks envisions these devices as being embedded in many different computing appliances, with the communications and microprocessor elements not particularly discernible to the user.
To implement such a concept, the existing and potential wireless infrastructure owner/operators must be involved in the evolution of the system together with wireless networking equipment manufacturers. These parties must have a reasonably common objective.
Forum for Development of Wireless Infrastructure
Establish a forum for those private- and public-sector entities involved in infrastructure for wireless data. Encourage analysis of the Spread ALOHA architecture and the establishment of strategic relationships among the parties, assuming that the effectiveness of that architecture is demonstrated. In conjunction with existing private-sector infrastructure providers, develop an implementation approach to overlay a Spread ALOHA architecture on existing wireless networks and determine the most appropriate frequency allocation. Coordinate and make appropriate filings with the Federal Communications Commission for the selected frequency use. This infrastructure implementation will be the critical factor in realizing such a nomadic, tetherless computing network.
PCS Data Standards
Encourage standardization proceedings to be initiated for a Spread ALOHA wireless air interface for PCS data and voice/data applications. The lowest-cost user PCMCIA card approach would require a hub infrastructure similar in coverage and spacing to voice PCS. Though it is recognized that the initial PCS implementations will be oriented toward telephony, the second implementation should be more attentive to data, thus offering a good opportunity for a broadband approachSpread ALOHA.
Integrate Tetherless Approach with NII Planning
The concept of tetherless data access to the NII should be integrated into other NII studies and planning. Most such studies and planning envision a person sitting at a desk. The tetherless concept is an important aspect of making full use of the NII in life beyond relatively static libraries, schools, and offices.
Abramson, Norman (editor). 1993. Multiple Access Communications: Foundations for Emerging Technologies, IEEE Press, New York.
Abramson, Norman. 1994. "Multiple Access in Wireless Digital Networks," invited paper, Proceedings of the IEEE, September.
1. Mobile Data Report. 1995. "BIS Estimates U.S. Market to Reach 9.9 Million Users in 2000," Vol. 7, No. 8, April 24.
2. Nelson, Paul, and Dan Merriman, BIS Strategic Decisions. 1994. "Wirelessly Enabling Portable Computers: A Major Growth Opportunity," The Red Herring, September/October, pp. 64–65.
3. Flynn, Laurie. 1994. "The Executive Computer: 3 Ways to Be Unplugged Right Now," New York Times, December 4.
4. Leibowitz, Dennis, Eric Buck, Timothy Weller, and John Whittier. 1995. The Wireless Communications Industry. Donaldson, Lufkin & Jenrette, New York, Winter 1994–95, p. 34.
5. Mobile Data Report. 1995. "McCaw Prices CDPD as Low as 8 Cents/K to Cover 75 Percent of its Markets in 1995," Vol. 7, No. 8, April 24.