5
The MPAR Concept

The ability of a phased array radar to form and steer the radar beam electronically, and to reconfigure the beam between any two transmitted pulses or even between transmit and receive modes, permits multiple functions to be carried out with the same radar. For example, the radar could direct narrow beams successively over a sector of the atmosphere to detect weather activity, a process that might take a minute or more depending upon the size of the sector and the dwell time required to measure the necessary weather variables in each beam direction. Interspersed with those observations, the radar could also search the approach path to an airport every few seconds to locate incoming aircraft. Thus the basic concept underlying the Multifunction Phased Array Radar (MPAR) initiative is to use a single radar type to carry out the multiple functions of weather and aircraft surveillance. The same radar might also carry out aircraft tracking missions, as well as possibly others to be determined (Figure 5.1).

FIGURE 5.1. Capabilities of agile-beam phased array radar are shown in a panoramic view. Illustrated are (a) surveillance scan through the planetary boundary layer (extending to 2 km) for mapping winds, (b) surveillance scan through a cumulus “Cu” cloud, (c) surveillance scan through a supercell storm, (d) high-resolution scan with a longer dwell time through the region in the supercell where the potential for tornado development exists, (e) scan that grazes the mountain contour for “surgical precision” avoidance of ground clutter, (f) determination of propagation condition, i.e., cumulative humidity along the beam between radar and the edge of the mountain, and (g) detection and tracking aircraft including noncooperating aircraft. Source: Zrnic, 2007, Reprinted with permission from AMS.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 26
5 The MPAR Concept The ability of a phased array radar to form and steer the radar beam electronically, and to reconfigure the beam between any two transmitted pulses or even between transmit and receive modes, permits multiple functions to be carried out with the same radar. For example, the radar could direct narrow beams successively over a sector of the atmosphere to detect weather activity, a process that might take a minute or more depending upon the size of the sector and the dwell time required to measure the necessary weather variables in each beam direction. Interspersed with those observations, the radar could also search the approach path to an airport every few seconds to locate incoming aircraft. Thus the basic concept underlying the Multifunction Phased Array Radar (MPAR) initiative is to use a single radar type to carry out the multiple functions of weather and aircraft surveillance. The same radar might also carry out aircraft tracking missions, as well as possibly others to be determined (Figure 5.1). FIGURE 5.1. Capabilities of agile-beam phased array radar are shown in a panoramic view. Illustrated are (a) surveillance scan through the planetary boundary layer (extending to 2 km) for mapping winds, (b) surveillance scan through a cumulus “Cu” cloud, (c) surveillance scan through a supercell storm, (d) high-resolution scan with a longer dwell time through the region in the supercell where the potential for tornado development exists, (e) scan that grazes the mountain contour for “surgical precision” avoidance of ground clutter, (f) determination of propagation condition, i.e., cumulative humidity along the beam between radar and the edge of the mountain, and (g) detection and tracking aircraft including noncooperating aircraft. Source: Zrnic, 2007, Reprinted with permission from AMS. 26

OCR for page 26
THE MPAR CONCEPT 27 Some military phased-array radars have been designed to perform multiple functions. For example, the AN/APG-81 and MP-RTIP airborne radars and the AN/SPY- 1 and SPY-3 shipboard radars routinely perform multiple functions, typically including a mix of surveillance and tracking activities and in some cases other functions as well. These functions are usually carried out sequentially using a prioritized control scheme, but the SPY-1 has even demonstrated concurrent weather and aircraft surveillance capabilities. One face of a SPY-1 system is used in the National Weather Radar Testbed (NWRT) facility in Oklahoma. There has been previous interest in the MPAR concept for aircraft and weather surveillance. A Federal Aviation Administration (FAA)-sponsored study (ITT 1997) determined that phased-array radar could meet most of the requirements for both aircraft and weather surveillance in an airport terminal area. However, the anticipated cost of such a system (in the mid-1990s time frame) was too high to warrant implementation of the concept at that time. Cost remains a major consideration in the feasibility of implementing the MPAR approach. The potential introduction of MPAR presents an opportunity to combine the diverse radar missions of weather surveillance, civil aircraft tracking and possibly homeland defense against airborne threats on a single1 standardized advanced technology platform. Implementation of an MPAR system to provide multiple functions could obviate, or at least diminish, the need for separate radar systems to support the individual functions. This could permit reduction in the total number of different radar types and radar units required to meet the nation’s coverage goals for weather and aircraft surveillance. Weber et al. (2007) provide a hypothetical example of this, wherein some 334 MPAR radars of one basic type (two distinct configurations) might replace 510 existing radars of seven unique types, while providing essentially the same coverage of weather and aircraft targets at 5,000 ft or more above ground level. If one assumed that the ongoing support costs per unit remained the same as the average for the systems the MPARs replaced, this would substantially reduce the annual system support costs. With the absence, in a full Active Electronically Steered Arrays (AESA) system, of a single high-power transmitter and a rotating antenna (both sources of major maintenance costs with many present-day radar systems), there is expectation that the ongoing support costs per MPAR unit should even be smaller. Furthermore, the support functions would be required to deal only with the one (or two) MPAR system types, in contrast to the seven or eight different systems providing those coverages today. That suggests important potential savings in engineering, logistics, and training areas. Developing time budgets for sequential allocation of scan capabilities to achieve the needed combinations of aircraft and weather data outputs could prove a difficult challenge. Consequently, two or more essentially separate radar systems, perhaps operating at different but nearby frequencies and using the same antenna, may be required to accomplish the desired concurrent missions (Weber et al., 2007). Operation in the S-band frequency range (2.7-2.9 GHz) of the current NEXRAD and FAA terminal area aircraft surveillance radars would provide desired characteristics 1 The JAG/PARP report (Chapter 6) and more recent supporting literature advocates the development of two separate but related radar designs, the MPAR and the Terminal-MPAR or T- MPAR. This distinction is discussed further below.

OCR for page 26
28 EVALUATION OF THE MPAR PLANNING PROCESS of sensitivity to both aircraft and precipitation targets, minimal effects of atmospheric attenuation along the beam path, and manageable range-Doppler ambiguity problems. In general, adequate sensitivity to precipitation echoes of interest would also provide adequate sensitivity to aircraft targets. Thus data on both target types could be extracted from each look in any beam direction; using polarimetric techniques would improve discrimination of aircraft from weather targets. To be sure, differing requirements on update rates might produce some redundant weather data in directions where frequent aircraft surveillance data are required. The work accomplished with the phased array radar (PAR) at the NWRT to help evaluate these possibilities has been excellent from both engineering and applications perspectives. The committee heard presentations from government and private sector representatives who stated strongly that the technology exists today to accomplish this vision. The cost arguments given are persuasive, namely that solid-state S-band transmitter/receiver modules are already being mass produced for commercial purposes and that further integration will lead to dramatic cost reductions for future radar applications. TECHNICAL CHALLENGES While the concept of a single, agile-beam multifunction radar design is appealing, many technical issues and questions remain to be resolved. True multifunctionality of PAR, with capabilities equivalent to those now available or soon to be available on the existing radar systems, has yet to be demonstrated. PARs have been used effectively for military purposes, and research at the NWRT has demonstrated that PAR technology can be used effectively for weather surveillance. However, not all of the needed weather capabilities have been demonstrated, nor has simultaneous aircraft tracking been demonstrated. Mark Weber’s presentation to the committee, illustrating how systematic electronic scanning with multiple radar frequencies might be used for both aircraft and weather surveillance purposes, did demonstrate that considerable thought has already gone into the issue of multifunction electronic scanning. Adaptive scanning, on the other hand, is an issue that remains to be addressed. The need for polarization diversity with the capabilities required for weather observations may present technical obstacles for phased array implementations. The NEXRAD polarization diversity upgrade will feature simultaneous transmission of orthogonal linear polarizations. It is not evident that this will be feasible with phased arrays, and sequentially alternating transmission of the two polarizations (with the attendant doubling of the dwell time requirement) may be needed. This issue is currently under study by Lincoln Laboratory (Weber et al., 2007). Another fundamental issue is the extent to which polarization diversity measurements will be degraded in beam directions other than in the principal horizontal and vertical planes, and whether these effects can be compensated through appropriate data processing (Zhang et al., 2008). Accurate and reproducible calibration of weather radars is essential for reliable interpretation of radar echoes. With phased array antennas the beam patterns and gain, as well as the polarization characteristics, change as functions of beam pointing angles. Moreover, the transmit and receive beam patterns may differ. These effects may be

OCR for page 26
THE MPAR CONCEPT 29 calculable for ideal conditions, but for actual systems the results must be experimentally verified and the associated errors quantified. With conventional single-transmitter radars the transmitted power and the receiver response characteristics can be readily measured. With an active phased array system this seems not to be possible, so calibration procedures will probably need to rely primarily on observing targets with known radar cross section. Such measurements would be required for all pointing angles, for both polarizations, and under various T/R module failure scenarios—not a simple task. Solar calibrations, possible only for a limited range of pointing angles, can provide some measurements of receiving antenna gain and beamwidth. The volume of weather data from an MPAR will be considerably greater than that from the current NEXRADs. More rapid volume scans and possible over-sampling in range and azimuth, in order to achieve higher resolution, will lead to data rates an order of magnitude higher than those experienced today. Undoubtedly the state of the art of signal processing hardware will allow sufficient speed to meet the need. However, communications, data storage, and data access may present substantial challenges. Secure historical archiving of radar track files may become a requirement. In designing the MPAR system, the data rates and volumes must be accommodated for all users. It was many years after the introduction of the NEXRAD system before the real-time data access problem had been satisfactorily addressed through CRAFT (Collaborative Radar Acquisition Field Test). In addition, easy access to archived data will be essential for system evaluation and product improvement research. A lesson learned with NEXRAD was that its initial proprietary architecture created obstacles to system upgrades. Open architectures have subsequently been adopted to facilitate the introduction of new capabilities and forecasting products. In an MPAR era, open architectures should be adopted to the greatest extent possible. COST ISSUES Chapters 5 and 6 and Appendices C and D of the Joint Action Group/Phased Array Radar Project (JAG/PARP) report provide information related to the anticipated cost of MPAR. Significant questions pertaining to developing realistic cost estimates for MPAR seem evident. Antenna element cost, overall system cost including requisite software, costs to site the MPARs nationwide, and costs of data dissemination and archiving are among the main uncertainties. Costs related to implementing the requisite capability for polarimetric observations and radar calibration need to be determined. Also, costs to conduct the adaptive scanning activities of an MPAR system are likely to be significant. In addition, costs for maintaining and adapting extensive software libraries after the prototype is deployed should be detailed. A complete MPAR cost analysis, including supporting data, an analysis of cost risk, and also a process for periodically revisiting and refining cost estimates, will be necessary to support any eventual implementation decision. As with any system designed to serve the needs of multiple federal agencies with multiple missions, MPAR cost increases are likely once the uses of MPAR by these various agencies are fully explored and understood. It is also quite possible that additional uses of MPAR will be identified prior to its operational phase, again with the

OCR for page 26
30 EVALUATION OF THE MPAR PLANNING PROCESS potential for increasing overall costs. These issues alone dictate the need for a flexible cost structure and a detailed cost uncertainty analysis. Furthermore, a mechanism is needed for revisiting the projected cost as the risk reduction project grows nearer to completion. Several significant cost elements in addition to the cost of the T/R modules, such as software development, system integration, training, development of data exploitation tools, and data dissemination and archiving, must be considered in the overall MPAR system cost estimates. The advent of multifunction phased array radar will add to the complexity of all of these elements. Many of these costs are likely to be non-recurring development costs that would be amortized over the total production run of MPAR systems. However, these costs will likely be front-loaded at the early stages of the MPAR program, including initial deployment, and will have a significant effect on required program funding profiles. Some cost elements, such as software development and training, will continue to some degree over the life of the program. All will need to be estimated with care by alternative and independent methods. Even well-managed software development projects are seldom completed on schedule, and their cost will inevitably multiply significantly as their completion date is delayed. The complexity of the software development effort needed for a complex multi- function phased array radar, especially if it includes adaptive scanning capability, should not be understated. Siting costs of MPAR will hopefully be minimized by employing existing legacy radar sites. It may be impossible to avoid the cost of both physical and Radio Frequency environmental impact studies, particularly if new radar frequencies and scanning plans differ significantly from the systems already in place. Sites near densely populated areas will require careful attention and may be subject to rollout delays through local opposition (due to possible concerns about appearance or (RF safety), adding to cost. Since each site is unique, it is unlikely that siting cost will drop as the deployment of new MPAR systems approaches completion. Organized opposition and public litigation, whether warranted or unwarranted, to a large scale radar deployment, as has been experienced in the cellular industry during tower buildout, is a possibility. Any necessary addition of site-specific software lockouts at low elevation angles would also add to deployment cost. Cost Models and Cost Risk Analysis Cost models and cost analysis packages are applied in both the government and private sectors, and MPAR costs will need to be modeled using accepted tools and practices. The Department of Defense (DOD) and the National Aeronautics and Space Administration (NASA) routinely use sophisticated cost estimation techniques employing statistical analysis to determine a range of likely costs for individual systems and entire programs. The profession of cost estimation, encompassing the disciplines of mathematics, statistics, application software and accounting, is active within industry and academia. The cost estimating methods used in the JAG/PARP report are highly preliminary and should be revisited in greater detail using professionally recognized methods.

OCR for page 26
THE MPAR CONCEPT 31 The DOD and other government agencies have recognized for years that both government and contractor cost estimates on major acquisitions are generally inaccurate and, despite best efforts, they are frequently too low (Anderson 2008). Examples of programs that have been put at risk because of (among other factors) poorly executed or overly optimistic cost estimates include: The NOAA GOES-R system, which, originally estimated to cost $6.2 Billion, was re-estimated to cost $11.4 Billion in May 2006 for a satellite scheduled for launch in 2012 (GAO, 2006a). This follows on the heels of the National Polar-orbiting Operational Environmental Satellite System (NPOESS) that is estimated to be approximately $3 Billion over budget and three years behind schedule. Software acquisition for Air Traffic Control (ATC) modernization at the FAA, which resulted in a 1997 Government Accountability Office (GAO) report that noted “many FAA failures in meeting ATC projects’ cost, schedule and performance goals, largely because of software related problems…” A decade later, the GAO found the FAA in an improved position on software acquisition, while ATC modernization is still listed as high risk. Cost estimates for the Next Generation Air Transportation System (NextGen) are currently at $15 Billion per year in 2005 dollars (GAO, 2006b). How accurate this estimate is and whether appropriate cost risk analysis has been applied in its estimation will be revealed over time. Point design cost estimates are highly unreliable, since component and activity costs (often referred to as Work Breakdown Structure or WBS elements) are probabilistic in nature (Anderson and Cherwonik, 1997). A total “best guess” cost based on an arithmetic sum of “best guesses” is almost guaranteed to be wrong. The cost estimation community has developed statistical techniques that account for both parametric relationships in costs and the uncertainty and risk in estimating them. The resulting cost estimates arrived at by Monte Carlo simulation can show a much higher cost risk than might be otherwise assumed, and total cost probability curves are often (and unfortunately) skewed asymmetrically to the right of average or “most likely cost” in the form of a lognormal distribution (Book, 2001). Figure 5.2 shows an example of a lognormal cost distribution. FIGURE 5.2. A lognormal probability density function. Source: Timothy P. Anderson, MCR Federal, LLC, Copyright 2004.

OCR for page 26
32 EVALUATION OF THE MPAR PLANNING PROCESS The cost estimates will also change over time due to unforeseen setbacks, changes in scope, and other problems that inevitably occur in the life of a project. An additional and frequent reason for cost growth is program underfunding or “stretch out”. This almost always leads to cost growth due to the continuation of fixed costs, additional inflation, and increased risk of programmatic disruption. CAPITAL ASSET PLANNING Successful completion of preliminary MPAR risk reduction R&D activity would enable the planning and budgeting phase for eventual deployment to begin. A large public acquisition such as an operational MPAR system would require preparation of detailed capital asset and business plans. A useful planning reference framework is described in the Office of Management and Budget (OMB) document “Capital Planning Guide: Planning, Budgeting and Acquisition of Capital Assets” (2006). In advance of any budget submission to OMB, written justification in the form of a “Capital Asset Plan and Business Case Summary” or “Exhibit 300” document is required by federal statute from all agencies of the Executive Branch. The Office of the Federal Coordinator for Meteorology (OCFM) with the Joint Agency Working Group (JAWG) in consultation with OMB and Congress would agree whether or not to fund the MPAR program to a level necessary to accomplish full-scale development and acquisition. ACQUISITION PLANNING AND CONTRACTING After completion of the planning and budgeting phase, the JAWG (or a newly created MPAR Joint Program Office) would proceed with development of the MPAR acquisition plan. When funding has been approved, the acquisition phase would begin. When requests for proposals for MPAR are drafted, a performance/capabilities based requirements strategy that leaves specific implementation decisions up to the contractor will offer more opportunities for creative and cost-effective solutions, as opposed to driving a hard requirement leading to a single architectural solution. Actions to upgrade or replace the nation’s surveillance radar systems would likely be phased in over time,2 and careful consideration of the acquisition methods and tempo will be needed. The Capital Programming Guide encourages “Modular Contracting,” which divides large acquisitions into smaller, more manageable segments or modules. Project phases could be executed and contracts could be written to optimally match achievable economies of scale for mass-produced components such as T/R modules, but also to provide budgetary and programmatic continuity to nurture and maintain a healthy contractor-industrial base for the life of the program. The Guide encourages robust competition at all stages of the acquisition, such as “competitive prototyping” before proceeding to full-scale development. In competitive 2 Some cost implications of different rates of phase-in are considered in the cumulative cost graphs presented in Appendix C of the JAG/PARP report.

OCR for page 26
THE MPAR CONCEPT 33 prototyping, both contractors and the government agencies can exchange information so that a common set of standards can be written into the requirements for a full-scale development. Awarding more than one development contract provides an incentive to competing contractor teams to strive for an optimum balance between performance and cost. An essential attribute of successful cost control in large acquisitions such as an operational MPAR network is the provision of sufficient and predictable project funding by sponsors at the correct time that is fully consistent with project plans. Interruptions, delays or reductions in the planned funding profile will cause significant and unrecoverable cost growth in a full-scale acquisition as well as in the R&D program. COMPARISON OF ALTERNATIVES Part of the Exhibit 300 submission is a section entitled “Alternatives Analysis.” In this section the desired capital expenditure is examined along with the status quo or current baseline, as well as alternate implementations including the one involving MPAR and T-MPAR. Other possibilities would include a next generation weather radar (that might or might not use phased-array technology) or a Center for Collaborative Adaptive Sensing of the Atmosphere (CASA) radar network, with later generational developments of legacy systems such as the ARSR-4, ASR-11, and the Terminal Doppler Weather Radar (TDWR), and various potential blends of candidate architectures. Such an alternatives analysis would of necessity consider solutions other than MPAR. Characteristics of Success and Failure in Large Government-Funded Engineering Projects The MPAR planning process should take note of best practices employed in the planning and execution of other successful megaprojects and adopt them as part of any MPAR planning and acquisition activity. In 1999, a National Research Council (NRC) study on Project Management at the Department of Energy DOE was convened to determine the root causes in long delays and budget overruns by as much as 50 percent for certain classes of projects over comparable projects at other government agencies or in the private sector (NRC, 1999). The Executive Summary of that report notes that DOE projects were abandoned before completion, cut back or delayed such that upon completion they served no useful purpose. The NRC report identified severe deficiencies in project planning and management at DOE in 16 areas. In Appendix C of the report, entitled “Characteristics of Successful Megaprojects or Systems Acquisitions,” a checklist was developed of conditions essential to, important to, and beneficial to success in a megaproject from the standpoint of general conditions, special conditions, and technical conditions of scope, costs and schedule for major stakeholders and participants in the project. MPAR planners, stakeholders and participants would benefit from applying the lessons learned from prior project successes and failures to avoid repeating the mistakes of the past.