Sharing Trunked Public Safety Radio Systems Among Federal, State, and Local Organizations
Federal, State, and local land mobile radio systems are generally single-channel analog FM voice systems, owned and operated by single agencies to perform a single well-defined mission. In many instances, it appears that technology and the communications organization have become entwined over the years, as part of a vertically-integrated radio communications organization. In this mode, there may appear to the organization little advantage to moving to more efficient radio systems, because the traditional radio best fits the traditional dispatcher organization.
However, there is currently considerable pressure for change for many Federal organizations because radio systems are becoming out-dated, without the prospect of sufficient resources to build a state-of-the-art communications system. In the State and local governments, the continued expansion of traditional radio systems has presented a crisis in spectrum crowding and has hindered the use of additional technological aids to efficient operation. The public safety community is particularly concerned about spectrum crowding, and Congress has instructed the FCC to consider additional spectrum for public safety uses.
Although the public safety community is possibly correct in their assertion that they are critically short of frequencies, it is also true that they have been slow to adopt new technologies that could provide much more efficient use of the spectrum. Thus, most public safety radio systems remain based on 50-year old spectrum technology — i.e., single-channel, 15 kHz bandwidth analog FM radio — which has been superseded in many services by more efficient technology. Possibly, the use of spectrum-efficient systems would allow public safety communications requirements to be met using the spectrum that they already possess. The public safety community has claimed that these more advanced systems are not suitable for public safety functions and have declared that public safety needs can be met only by massive amounts of new spectrum. This chapter will examine shared, multisite, trunked radio systems to see whether it might be suitable for general government operations (including public safety) as an alternative to supplying massive amounts of additional spectrum.
Talk Groups. Trunked systems can establish a communication channel to individual radios or to groups of radios. The groups of radios are called "talk groups". A given radio can choose which of the talk groups it listens to at any given time. Similarly, a given speaker can choose which talk groups are the recipient of his call. These talk groups can be rapidly modified under a number of options. In a time of emergency, for example, a combination of agencies providing emergency services could be linked together in a single new talk group. Radio users not belonging to a given talk group would be unaware of any activity on that talk group, except for the possible consequence of general crowding on the system (which we have already shown would be very small).
The talk group process limits user interaction to only those interactions deliberately set up by talk group management. This process can be further enhanced through encryption and priority access. Encryption systems used with some digital trunked systems are sufficiently secure that they meet NSA requirements for classified information transfer, which are far more secure than traditional methods of "scrambling" voice channels for privacy. Systems using digital vocoders scramble the transmitted bits by multiplying them with a long binary code key and performing the reverse multiplication in the receiver. Some trunked systems allow over-the-air re-keying of radios.
Multiagency Talk Groups. A major shortcoming of independent radio systems is that there are important times when the independent agencies need to provide a joint response to a common task, i.e., they need to communicate with each other via radio. Therefore, in addition to each agency having its own set of independent channels, there is often an additional set of shared channels that can be used when circumstances require cooperative work between agencies. The public safety "mutual aid" channels are examples of shared channels intended to temporarily link multiple agencies. When not being used for shared purposes, these control channels remain unused (i.e., wasted) even though base station transmitters and many mobile/portable radios have been equipped at considerable expense to use these channels.
When the multiple agencies are all on the same shared system, the function of communicating between agencies during emergencies or other shared projects becomes trivially easy. A talk group is established between all of the radios that need to participate. All members of the talk group can instantly communicate, using the standard radios that they use for all functions. Since this is a standard system talk group, all of the regular system features are available to it, including encryption and priorities, wide-area coverage (assuming that the trunked system is capable of wide area coverage), telephone interconnect, digital messages, etc.
Multisite systems. Trunked radio systems can be extended to multiple sites when service is needed over a larger area than can be served by a single site. Multisite systems can be built in two considerably different ways, the "basic" way and the "enhanced" way. The basic way merely involves the construction of multiple trunked repeater sites. The sites are not interconnected, except coincidentally through the telephone system. A user keeps a list of which frequencies are available at each site, as well as the approximate coverage area of each site. Depending on the user's location, the user selects the frequency appropriate for the nearest site and communicates with other users being serviced by that site. Alternatively, the user may contact the nearest site to interconnect to the local phone system and complete his conversation via a local or long distance telephone circuit. The process of completing a call to a mobile user can be more frustrating, particularly if the location of the mobile user is not known, since the caller will not know which site to use. The caller might try calling each site, to test whether the desired radio was within range of that site. Eventually, the caller will obtain a response when the correct site is used. The basic multisite system works grudgingly to give wide-area coverage, but only with considerable effort and persistence on the part of the users. Most older existing multisite systems are of the basic type.
An enhanced multisite system imposes a much greater level of system complexity, but provides a much higher level of functionality and user convenience. The enhanced system knows the location of every user in sufficient detail to chose which of the multiple sites should be used to contact a given user. If quick response is needed to contact a talk group, the system must know this information in advance, rather than going through a time-consuming automatic poll of talk group members at all sites. The details of obtaining user location information differs between systems. Some systems continually and automatically test which mobile units are within range of each of the sites. This data is stored in a central database, which is accessed whenever a call needs to be placed to a mobile user. In addition to an extensive control and data systems that coordinate the action of the individual sites, a voice-channel switching system is needed to route communications between the sites. The total cost of a complex multisite system is several times as expensive as the simple multisite system. In addition, the complex multisite system will require much greater system database maintenance, intersite communications, and general cost.
The level of service and the cost of the two types of multisite systems is greatly different. In the remainder of this paper, we will use the term "multisite system" to refer to the enhanced multisite system, unless the term is specifically modified with a "basic" descriptor.
Spectrum efficiency. Besides extending the total coverage area of a trunked system, the use of multiple sites provides a mechanism to greatly increase the amount of service available from a given number of channels. The coverage area of a given site can be deliberately reduced by using less transmitter power, lower antenna gain, lower antenna towers, and the use of directional antennas. Although the coverage area of each site is reduced, this technique allows a given frequency to be re-used in another site. Cellular systems, for example, typically reuse frequencies every 12 sites. Smaller coverage areas means an increased number of sites and an increased amount of frequency re-use in a given geographical area. Reduced-coverage sites are usually built only in large metropolitan areas, where the density of customers provides more traffic than can be serviced with full-size sites. The ability to design various size cells provides great flexibility in the types of service provided to large buildings, stadiums, and other large concentrations of traffic. Because small coverage areas mean that a given mobile conversation moves from one site to another relatively often, the use of small coverage areas is only practical where switching between sites is done automatically and with a minimum amount of interruption to the user.
The significance of multiple site technology is that frequency crowding can be greatly reduced in larger metropolitan areas, which is precisely where acute frequency crowding is a problem. The major disadvantage of multiple site architectures is increased construction and operational cost of the large number of sites.
At present the typical situation for individual agencies is to have certain geographical areas where coverage is required (because of the presence of a large number of important users), as well as larger area where coverage is desirable. In many cases, the agency communications budget can only afford to cover the most important areas. Each agency will have somewhat different areas of required coverage. The large multisite trunked system will provide coverage that is at least the sum of the required coverage areas. The total coverage area will generally be much larger than individual agencies could afford on an individual basis, including the areas of desired coverage in addition to the required coverage areas. Moreover, the multisite system architecture makes it convenient to operate in many geographical areas without specialized knowledge of frequencies, and the architecture allows easy deployment of new sites when it becomes desirable to do so.
The public cellular telephone system is an example of a multisite trunked radio system. The need to go through the time-consuming processes of dialing a number and waiting for an answer on the other end makes cellular phones an example of a process that has not been optimized to give a quick response. It is also an example of trunked radio systems that do not offer encryption or priority response (at least not with the present AMPS standard). It should not be used as an example of a trunked radio system that would be proposed for public safety. However, some features of cellular telephones illustrate capabilities that would be very useful for public safety, including nationwide access with a common radio, very high spectrum efficiency (serving 25 million users with 50 MHz of spectrum), and the very economical mass production of radios via a common technical standard serving a mass market.
Sharing. Trunked single-site or multisite systems can be shared among a mix of users, with each type of user having an appropriate set of talk groups and priorities. At a first level of approximation, one talk group should not even be aware of the existence of other users. Nevertheless, the presence of other users becomes important when the system becomes crowded or when priorities are invoked. There are several philosophies about with whom one would be willing to share a trunked radio system. The traditional answer is to share a trunked radio system only with users who are very similar to you. (Thus, the FBI might be willing to share with the Secret Service, but would not be willing to share with a pizza delivery service.) This philosophy is prevalent because similar users are likely to share much of the same "agency culture." Among similar users it will be relatively easy to decide on operating procedures, chains of command, situations where priorities need to be invoked, special features that should be available on the system, etc. Unfortunately, similar users are likely to experience emergencies at the same time. If all agencies sharing a system each experience an emergency at the same time, there will surely be a massive channel shortage at that time. Moreover, each agency will simultaneously have a legitimate reason to invoke high priorities at the same time, meaning that priorities will not be effective in reducing crowding by eliminating lower priority traffic. During a general emergency, however, every user will be high-priority, and no one will benefit from the "high priority" label.
An alternate option is to share with dissimilar agencies whose major traffic peaks and emergencies are likely to occur at different times from your own. Thus, the "National Drought Relief Agency" could ideally share a system with the "Bureau of Flood Control." Of course, it may be difficult to find exact opposite emergency requirements, but the principle works to some degree whenever different agencies (or different functions within a single agency) share a system. The point is that diversity in traffic demand provides a major advantage in shared trunked systems.
Another alternative would be to mix many low-priority users with high priority users. During emergencies, the low priority users will see their access greatly limited, while their channels become available for high-priority functions. In cases where a single agency has multiple functions, the radios themselves might be moved from low-priority to high-priority functions, depending on the needs. The number of channels can be sized on the basis of the large amount of daily traffic from the low priority users. If the large number of low-priority users cause a 10-channel system to be built instead of a 5-channel system, the high priority users will have twice as many channels to pre-empt when they are needed.
A problem with adding too many low-priority users is that the low priority users may not want to pay for the features (encryption and key management, priority access, better coverage in remote areas and within buildings, additional system robustness, etc.) needed by the high priority users. Similarly, a great diversity of users may result in the system having too many special features (a special feature for each special-purpose need of a user). It may be difficult to entice the low-priority users to join the system; lower fees may be appropriate for lower priority users.
System Operation: Dispatch/LMR. The multisite, trunked, radio system provides radio service using techniques somewhat different from the traditional dispatch operation. First, the regular dispatch channel (e.g., "OPS 1") is not a specific frequency, but instead is a talk group. All users that expect to monitor OPS 1 will carry radios that include OPS 1 in their list of active talk groups. OPS 1, like any other talk group, may change frequencies several times during the day. If a transmitter channel needs to be shut down at a site, the talk groups simply change to another frequency. OPS 1 might include some users serviced by several different transmitter sites; each of those sites would probably be using a different frequency for OPS 1. Users would have no knowledge of what frequency was being used by OPS 1; they would only know that the OPS 1 calls were received on their radio. User calls would be placed to OPS 1, OPS 2, etc., by addressing the desired talk-group. The system keep tracks of which OPS 1 talk group members are nearest to each site and establishes the required connections.
System administration. If users from many different agencies share the same multisite trunked radio system, there will surely be many points of view on any particular aspect of system operation. One agency culture may favor tight operational controls with all traffic being recorded for possible future documentation and analysis. Another agency may need the latest in digital messaging and advanced computer database access. Another agency may want good quality telephone interconnect and ease of operation. Still another agency may need encryption and robust emergency response capabilities. Since these are prime areas of concern for the respective agencies, agency participation in the shared system may depend on seeing these needs met.
Various methods of deciding these questions can be suggested. A User's Council provides one source of guidance for system administration. The authority of the User's Council (by whatever name it is called) can vary from an informal advisory body to the major ruling organization that issues administrative fiats to a paid system operator. In many cases, the primary authority will come from a corporation (for-profit or non-profit), a State government body, a Federal agency, or a joint State/private partnership. These authorities would find an advisory User Council to serve a very useful planning and public relations function. At this time, it seems difficult to judge what specific organizational arrangements will be most effective.
Whatever administrative arrangements are effected, it is clear that the users will need to have assurances that their operational needs will be met, that critical information will be protected, that the system will be robust and flexible in challenging situations, that the cost of the system will be reasonable, and that future arrangements can be flexibly modified to meet new situations. The credibility of these assurances may be a problem, particularly for Federal participants whose culture suggests a higher level of security and robustness concerns than the State organizations that may be operating the system. (Whether this concern is actually justified is only part of the problem. The perception may prevent participation by several Federal agencies who could be mutually beneficial participants in shared systems.)
The preceding technology descriptions can be the basis for a new architecture for government LMR systems. This new architecture would be based on a multisite radio system shared among many different government agencies. It would replace the large number of independent government radio systems that currently service government functions on a Federal, State, and local basis. This single government radio entity would serve a metropolitan area or provide a statewide coverage. Federal users would be a part of the appropriate local or statewide system serving their area of operations. The individual systems should be built to uniform technical standards (possibly APCO-25) in such a way that Federal Government users could easily move between systems on a "roaming" basis, providing national coverage when required.
These systems could be set up on a state-by-state basis, as a Federal network (perhaps by a Justice/Treasury/FEMA effort), or by private corporations. We note that the mainline common carriers have been reluctant to upgrade cellular/PCS capabilities to serve the specialized government/public safety market. We believe that the cost/benefits ratio of such a system would encourage many government agencies to become part of such a system. Although the costs of a system would be great, the level of service would be much higher than with conventional radio systems. Indeed, unless this system (or some comparably advanced/expensive system) is constructed, government agencies will suffer from a crippling stagnation of communications-based operational efficiency. The cost of this system would be considerably less than if each agency were to build independent systems providing the required level of service. Many individual agencies and municipalities will buy into the system over the next 5-15 years as their existing radio systems are replaced for obsolescence or expanded to meet growth needs. Private and corporate use of the system could be allowed, when the need for a higher level of communications robustness justifies the cost (this would be almost inevitable if the service were provided by a private corporation).
The initial cost of such systems will be quite high. Present cost for an APCO-25 compliant radio is expected to be $2000-3000. A mass market based on open standards and competition between vendors is expected to drive costs down. (Note that cellular telephones cost $2000-3000 when they were first introduced.) The infrastructure for these systems will also be very expensive, since typical enhanced multisite trunked systems should be expected to cost 3-10 times more than basic multisite systems.
We believe that such an architecture will be viable for a minimum of 25 years. Communications progress is hard to foresee beyond that period, but the major threat to such shared systems will probably be the availability of commercial services that are equally competent and cheaper. Within the context of the system itself, we would expect a continuous process of adding additional functions and services, as technology allows and needs demand.
Examples of Wide-Area Shared Systems
The system described in the preceding paragraph has not yet been matched in fact. However, some systems have been proposed and are apparently moving towards implementation which seem to be driven by many of the same considerations as in the vision. We have included a brief description of a sampling of these systems for two reasons. First, they tend to validate the arguments describing a need for such shared systems. Second, they illustrate various approaches for achieving the fact of these systems. Possibly some synthesis of the approaches described here will offer the best of all organizations for these systems.
State of Colorado. The State of Colorado is planning a statewide digital trunked radio system (DTRS), based on APCO-25 standards. Planning began in 1991 within the Division of Telecommunications, studying advanced designs for an improved statewide system for the State Patrol. Eventually, components of the Highway Maintenance and Natural Resources Departments were included. Beginning in 1992, an extended series of information meetings were undertaken to gather public support of this system. A series of working committee meetings were set up to provide advice on the services and the administration of the system. These meetings included many public safety and municipal communication professionals, as well as industry, Federal agency, and FCC and NTIA personnel. This working committee issued a report in June 1995.[EN #1]
A 6-phase schedule of implementation has been proposed for DTRS, beginning with the Denver Metropolitan area in 1996. The system is expected to cost somewhere near $120 million, not counting the microwave backbone which is already mostly in place. It is anticipated that the system will be built using the 800 MHz public safety bands. The State legislature would be expected to provide much of the funding, with some expenses recovered from monthly fees and some reimbursed construction costs. The Division of Telecommunications will own and operate DTRS, but local governments and Federal agencies have been invited to participate. A strong user's group is expected to be set up to help govern the system.
The implementation model assumes that several types of users will want to associate themselves more-or-less closely with the State network. The State Patrol, Highway Maintenance, Natural Resources, and Corrections will be full members, based on legislative fact.
Client members will be given service on DTRS, in exchange for the monthly subscription fee. Client members are typically small communities that have traditionally not built their own radio systems.
Integrated members will forego building their own radio systems, but will contract with the Division of Telecommunications to design and build a system for them, which will become part of the DTRS. Integrated members will probably pay a monthly service fee, as well as the incremental cost of building their part of the system. Typical Integrated members include small-to-medium size towns, whose communications needs are too large to piggy-back on the unmodified DTRS capabilities, but which have chosen not to build their own facility.
Cooperating members will build their own radio facilities to meet their own needs, but will design it to become part of the Colorado network. Cooperating members might include city/counties with a large population who are currently operating an extensive radio system. The Cooperating network might replace the DTRS Network in areas where the Cooperating network provides coverage. In exchange for providing services to all members of the Colorado network, the cooperating members might receive payment for services, as well as roaming rights on the remainder of the Colorado Network.
Associated members, like Cooperating members, build their own systems, but do not fully integrate them into the DTRS. Sharing between the Associated network and the DTRS is on a voluntary and limited basis, though an Associated network is fully capable of operating as part of the DTRS. An Air Force base might become an Associated member, maintaining full control over its own system, though finding it convenient to share limited roaming privileges with DTRS members. Some municipalities may initially join the DTRS as Associated members, using this status as a halfway point while deciding whether to become Cooperating members.
Finally, a limited number of Commercial members might be allowed, especially in areas of the state where other communications were not available. Commercial members might include guides in wilderness areas or ranchers in remote areas of the state. Commercial members would pay a monthly fee.
The Colorado DTRS is currently proceeding on schedule. Several municipalities are in conversation with the Division of Telecommunications regarding coordination of their municipal radio improvement plans with the DTRS.
The availability of a frequency band that could be used by Federal agencies and State agencies remains a problem, though the Division of Telecommunications has indicated that sufficient spectrum exists near 800 MHz to meet current needs. The FCC has indicated that the 800 MHz public safety bands could be used for this purpose.
State of Michigan. The State of Michigan is building a statewide digital trunked radio system based on Motorola Astro technology. The system will include 168 sites, including upper and lower peninsulas, with an estimated cost of $187 million. This will provide mobile coverage for 97% of the area of the state. Operation of the system is scheduled for October 1996. A total of 66 frequency pairs in the 821-824 MHz and 866-869 MHz band will be used in the system. The first implementation will include 1500 radios for State functions and 3000 radios for local government functions.
Users on this system include State Police, 911, and all other State public safety functions. Municipal governments are being invited to join. The City of Lansing (the State capitol) will be part of the system. Radio users will buy their own radio, pay a $250 entrance fee, and pay $300/year for service.
Some Federal law enforcement agencies have asked for some access to the system, though this is currently not intended to replace any existing Federal networks.
Racom (Iowa and Surrounding States). Racom is a commercial SMR company that has been supplying analog SMR services using 100 sites covering Iowa, much of southern Minnesota, and parts of Nebraska, South Dakota and Wisconsin. These older analog sites provide telephone interconnect and allow users operation on all sites in the system. Charges are $10/month plus $.25/minute airtime charges for telephone interconnect.
Recently Racom announced plans to build a multisite digital trunked radio system, eventually utilizing 200 sites over approximately the same multi-state geographical area. The new system will provide digital and analog voice and other digital services using Ericsson EDACS technology. The first part of the planned network is operational with 9 sites operating in Polk County (Des Moines area) which are providing services to the 300 radios of the Polk County Sheriffs Department. Cost to the Sheriffs Department is $15/month/radio plus $.30/minute airtime for telephone interconnection.
The digital trunked service includes both business and law enforcement customers. Law enforcement radios have "ruthless preemption" privileges and can immediately preempt business user channels if law enforcement needs another channel. The wide area system is targeted toward the "high priority" market, including private and government public safety operations, utilities, and similar customers. The frequencies for this operation come from the 800 MHz SMR frequencies. Although no encryption is in use at present, the Sheriffs department can encrypt transmissions whenever they feel that it is needed.
State of Louisiana. The State of Louisiana is converting a large number of independent radio systems to a single 125-site trunked system based on Motorola SmartZone analog technology. This system will provide coverage to 95% of the state and will include all State communications functions. Frequencies will come mainly from the 821-824 MHz and 866-869 MHz public safety bands. It is anticipated that some municipal governments will also want to coordinate their radio systems with this network. Non-State users will be asked to pay $200/year.
San Diego Navy Trunked Radio System. The Navy Public Works Center (NPWC) operates a trunked radio system for Federal agencies in the San Diego area. At present, about 650 radios are in operation on the 13-channel single-site trunked system, which uses frequencies in the Federal 406-420 MHz band. A total of 2400 radio users have asked for permission to be placed on the system, but funds to purchase new radios compatible with system operation have limited the use of the trunked system (cost: $1500-2500 apiece). NPWC charges $17/month/radio, including maintenance on user radios. Telephone interconnect costs $12/month more, but no airtime charges are assessed.
Most of the radio users are military operations, whose previous frequencies were in the 138 MHz, 162 MHz, and 406 MHz bands. Other users include civilian Federal agencies, especially law enforcement agencies. There is a limited capability to interconnect with 800 MHz San Diego City departments, based on console patches.
The NPWC mentioned considerable difficulty in obtaining additional channels in the 406 MHz band, apparently due to claims of heavy local demand by other agencies. RSMS channel occupancy measurements showed that the NPWC trunked system accounted for the majority of usage in the band.
1. State of Colorado, Division of Telecommunications, Digital Trunked Radio System: System Requirements and Operational Plan, June 1995.
Return to Table of Contents.