Controlling Satellite Communication System Unwanted Emissions in Congested RF Spectrum

Donald Olsen The Aerospace Corporation P.O. Box 92957 Los Angeles, CA 90009-2957

Roger Heymann National Environmental Satellite Service NOAA-NESDIS Building SSMC1, 5th Floor 1335 East-West Highway Silver Spring, MD 20910

ABSTRACT

The International Union (ITU), a United Nations (UN) agency, is the agency that, under an international treaty, sets spectrum usage regulations among member nations. Within the United States of America (USA), the organization that sets regulations, coordinates an application for use, and provides authorization for federal government/agency use of the radio frequency (RF) spectrum is the National and Information Administration (NTIA). In this regard, the NTIA defines which RF spectrum is available for federal government use in the USA, and how it is to be used. The NTIA is a component of the United States (U.S.) Department of Commerce of the federal government. The significance of ITU regulations is that ITU approval is required for U.S. federal government/agency permission to use the RF spectrum outside of U.S. boundaries. All member nations have signed a treaty to do so. U.S. federal regulations for federal use of the RF spectrum are found in the Manual of Regulations and Procedures for Federal Radio Frequency Management, and extracts of the manual are found in what is known as the Table of Frequency Allocations. Nonfederal government and private sector use of the RF spectrum within the U.S. is regulated by the Federal Communications Commission (FCC).

There is a need to control “unwanted emissions” (defined to include out-of-band emissions, which are those immediately adjacent to the necessary and allocated bandwidth, plus spurious emissions) to preclude interference to all other authorized users. This paper discusses the causes, effects, and mitigation of unwanted RF emissions to systems in adjacent spectra.

Digital are widely used in today’s satellite communications. Commercial communications sector standards are covered for the most part worldwide by Digital Video Broadcast - Satellite (DVB-S) and digital satellite news gathering (DSNG) evolutions and the second generation of DVB-S (DVB-S2) standard, developed by the European Telecommunications Standards Institute (ETSI). In the USA, the Advanced Systems Committee (ATSC) has adopted Europe’s DVB-S and DVB-S2 standards for satellite digital transmission. With today’s digital modulations, RF spectral side lobes can extend out many times the modulating frequency on either side of the carrier at excessive power levels unless filtered. Higher-order digital modulations include quadrature phase shift keying (QPSK), 8 PSK (8-ary phase shift keying), 16 APSK (also called 12-4 APSK (amplitude phase shift keying)), and 16 QAM (quadrature amplitude ); they are key for higher spectrum efficiency to enable higher data rate transmissions in limited available bandwidths. Nonlinear high-power amplifiers (HPAs) can regenerate frequency spectral side lobes on input-filtered digital modulations. The paper discusses technologies and techniques for controlling these spectral side lobes, such as the use of square root raised cosine (SRRC) filtering before or during the modulation process, HPA output power back-off (OPBO), and RF filters after the HPA. Spectral mask specifications are a common method of the NTIA and ITU to define spectral occupancy power limits. They are intended to reduce interference among RF spectrum users by limiting excessive radiation at frequencies beyond the regulatory allocated bandwidth.

The focus here is on the communication systems of U.S. government satellites used for space research, space operations, Earth exploration satellite services (EESS), meteorological satellite services (METSATS), and other government services. The 8025 to 8400 megahertz (MHz) X band can be used to illustrate the “unwanted emissions” issue. 8025 to 8400 MHz abuts the 8400 to 8450 MHz band allocated by the NTIA and ITU to space research for space-to-Earth transmissions such as receiving very weak Deep Space Network signals.

The views and ideas expressed in this paper are those of the authors and do not necessarily reflect those of The Aerospace Corporation or The National Oceanic and Atmospheric Administration (NOAA) and its National Environmental Satellite Service (NESDIS).

KEY WORDS: Unwanted emissions, RF spectrum, satellite communications, modulations, communications technology, transmitter linearization

INTRODUCTION AND BACKGROUND

There is a need for satellite communication systems using the RF spectrum to control “unwanted emissions,” which are defined to include out-of-band emissions (immediately adjacent band), plus spurious emissions to preclude interference to other authorized spectrum users. The ITU, through international treaty, sets standards for member nations and provides authorization for RF spectrum use beyond a nation’s boundaries. Member nations determine their own sub-allocations within the ITU regulations for their own domestic use. In the USA the NTIA regulates, coordinates, and provides authorization for agencies of the federal government to use the RF spectrum as required by federal law. Similarly, the Federal Communications Commission (FCC) regulates the nonfederal use of the . These organizations publish documents defining the permitted levels of both in-band and out- of-band transmitter power. The FCC and the NTIA work together in their respective domains to promote a coordinated use of the various bands within the structure of the ITU limits. The Space Frequency Coordination Group (SFCG), with input from the Consultative Committee for Space Data Systems (CCSDS), plays a significant role in advising the ITU, including proposing spectrum mask definitions to the ITU.

Traditionally there has not been a great emphasis by U.S. federal government agencies to use the RF spectrum as efficiently as they might. However, due to the increasing demand for spectrum over the last few years, the spectrum use regulators have greatly increased their emphasis on efficiency. In 2003, the Department of Commerce was directed by the White House to prepare recommendations to improve spectrum management. This became part of a presidential directive on using spectrum more efficiently, titled “Improving Spectrum Management for the 21st Century.”

From that paper, President George W. Bush signed a “Presidential Determination: Memorandum for the Heads of Executive Departments and Agencies.” He opens that memo, dated June 5, 2003, by stating:

The existing legal and policy framework for spectrum management has not kept pace with the dramatic changes in technology and spectrum use. Under the existing framework, the Federal Government generally reviews every change in spectrum use. This process is often slow and inflexible and can discourage the introduction of new technologies. Some spectrum users, including Government agencies, have argued that the existing spectrum process is insufficiently responsive to the need to protect current critical uses. 1

Later the President wrote,

1 Presidential Memo on Spectrum Policy, Office of the Press Secretary, June 5, 2003.

2 In May 2003, I established the Spectrum Policy Initiative to promote the development and implementation of a U.S. spectrum management policy for the 21st century. This initiative will foster economic growth; promote our national and homeland security; maintain U.S. global leadership in communications technology; and satisfy other vital U.S. needs in areas such as public safety, scientific research, Federal transportation infrastructure, and law enforcement. 2

Then in the same memo he directed:

the Secretary of Commerce to prepare recommendations for improving spectrum management. The Secretary of Commerce then established a Federal Government Spectrum Task Force and initiated a series of public meetings to address improvements in policies affecting spectrum use by the Federal Government, State, and local governments, and the private sector. The recommendations resulting from these activities were included in a two-part series of reports released by the Secretary of Commerce in June 2004, under the title Spectrum Policy for the 21st Century - The Presidents Spectrum Policy Initiative (Reports).

Within 1 year of the date of this memorandum, the heads of agencies selected by the Secretary of Commerce shall provide agency-specific strategic spectrum plans (agency plans) to the Secretary of Commerce that include: (1) spectrum requirements, including bandwidth and frequency location for future technologies or services; (2) the planned uses of new technologies or expanded services requiring spectrum over a period of time agreed to by the selected agencies; and (3) suggested spectrum efficient approaches to meeting identified spectrum requirements.3

This was followed up by a memo from the Honorable Secretary of Commerce Carlos M. Gutierrez to the Honorable Conrad C. Lautenbacher, Jr., the Under Secretary of Commerce for Oceans & Atmosphere within NOAA, where in part he stated that:

As directed by the Executive Memorandum, your agency’s plan must be submitted no later than November 30, 2005, and shall include:

(1) spectrum requirements, including bandwidth and frequency location for future technologies or services;

(2) the planned uses of new technologies or expanded services requiring spectrum over a period of time agreed to by the selected agencies; and

(3) suggested spectrum efficient approaches to meet identified spectrum requirements.4

With funding by NOAA-NESDIS, The Aerospace Corporation has investigated technologies related to spectrum conservation and bandwidth efficiency through analysis and development of a Geostationary Operational Environmental Satellite (GOES-R) radio communications test bed. This test bed was specifically tailored for the processed data uplink (PDU)/global rebroadcast (GRB) data and included an uplink signal processor, a satellite L-band downlink traveling wave tube amplifier (TWTA), downlink channel, and a downlink signal processor. The processing currently includes error correction encoding and modulation for the uplink and demodulation and decoding for the downlink. It presently does not include an uplink X-band transmitter, channel, or receiver and is connected directly to the L-band TWTA

2 USA Federal Government Presidential Determination: Memorandum for the Heads of Executive Departments and Agencies, 30 November 2004. 3 Ibid. 4 Letter on Spectral Efficiency from Secretary of Commerce, Carlos M. Gutierrez, to Under Secretary of Commerce, Conrad C. Lautenbacher, Jr., 10 March 2005.

3 input. The signal processors are very flexible and implemented in field-programmable gate array (FPGA) technology. The encoding and decoding are done with an AHA company turbo product encoder/decoder chip. The nomenclature for this chip is AHA4540A. We have built modulators and demodulators, including phase and symbol tracking for binary phase shift keying (BPSK), QPSK, and 16 APSK.

This paper discusses technologies and techniques with which future systems such as the emerging new-generation NOAA GOES-R satellite series in response to the above directive might more efficiently use the limited spectrum than did prior GOES systems and also deal with an essential related issue of controlling unwanted emissions. The technologies offer to significantly mitigate the increase in bandwidth necessary to accommodate the greatly increased data rate required for this and future systems. These include data compression, higher-order modulation formats, modern, more powerful, higher code rate forward error correction coding, and improved HPA linearization for a greatly reduced OPBO. This paper also discusses the causes, effects, and mitigation of unwanted RF emissions and increased radio frequency interference (RFI) on systems using adjacent spectrum. These technologies reduce the required frequency separation, called the frequency guard band, and thereby improve spectrum utilization.

ANALYSIS AND DISCUSSION

In the big picture view, bandwidth efficiency is the ratio of the precompression cumulative GOES-R sensor data rate to the required total RF bandwidth for transmission. Bandwidth efficiency is measured in bits per second per hertz of channel bandwidth. Today’s digitally modulated signal emissions can extend out to many times the necessary bandwidth on either side of the carrier and thus require a guard band between spectrum users (this can be seen later in the paper, in Figure 3, where “roll-off rate” is discussed). Since the power spectral density doesn’t fall off to zero immediately outside the main modulation spectral lobe and because there are usually many spectral side lobes, bandwidth efficiency must take into account the guard band required between the wanted signal’s spectral lobes and any adjacent signals. This guard band is necessary to reduce the mutual RFI.

Efficient use of RF spectrum, given a certain data compression factor, requires effective use of three features. The first is minimizing the necessary bandwidth, a frequency management term, the corner point on the spectral mask where the downward slope starts. The second is minimizing the guard band needed between adjacent spectrum users, and the third is limiting maximum power level of unwanted emissions (out-of-band plus spurious emissions). The latter is done by working to meet the recommended frequency masks on power limits issued by the NTIA and ITU and likely should include as well, in our view, negotiations with adjacent band users.

Minimizing necessary bandwidth must be done consistently within the system trade space dictated by allowable transmitter power and receiver sensitivity degradation from signal distortion within the limits of link geometry, data rate, and link availability in the propagation environments. Avoiding degradation of adjacent signal reception or limiting spectrum usage may require a steeper slope than that available with any existing regulatory mask. The width of the guard band is set by the proximity of other adjacent spectrum users or by the slope of the envelope of the required spectrum mask and by achievable filter technology. The smaller value governs the spectral width.

Figure 1 shows the applicable SFCG spectral masks.5 The steeper curve is for signals with symbol rates exceeding 2 megasymbols per second. It has a slope of about 80 dB () per decade. The shallower curve is for lower data rates and requires a slope of only about 43 dB per decade. They provide 16-to-1 and 6-to-1 bandwidth ratios, respectively, at the 60 dB down points. The NTIA has also defined spectral masks for many applications. In particular the mask to which space applications signals

5 Space Frequency Coordination Group (SFCG), Figure 1 of Recommendation 22-2R2, p. 3.

4 must adhere is shown in Figure 2.6 Here dBsd stands for dB of spectral density relative to the peak value. The necessary bandwidth is set such that the upper corners of the two-sided width of the main frequency spectral lobe of the signal are 8 dB down from the spectrum peak on the mask rather than having the corner point be at 0 dB, as in the SFCG mask. Furthermore the NTIA mask includes a slope of precisely 40 dB/decade of frequency offset. The reference frequency for this ratio is with respect to the single-sided, necessary half-bandwidth. Both masks continue outward and downward until the curves reach the –60 dB point with respect to the peak spectrum value. After that they remain flat at a –60 dB floor. The guard band for this NTIA mask is a factor of 19 times the half-necessary bandwidth for the lower mask corner, a significant factor in terms of 95% loss of spectrum efficiency. This slope is much less stringent than the 16-to-1 and 6-to-1 ratio values for the SFCG mask. Therefore the steeper curve of the SFCG mask provides much better bandwidth efficiency than the NTIA mask provides.

Figure 1. Spectral Emission Masks 0

-10 Rates > 2Mbps -20 Rates < 2 Mbps -30

-40

-50

Spectrum Attenuation Attenuation (dB) Spectrum -60

-70 0123456789 Frequency-off-Carrier to Symbol Rate Ratio (F/Rs)

Fourier analysis shows that the customary unfiltered square modulation pulses have a 2 [sin(πf/Rs)/(πf/Rs)] normalized power spectral density normalized to 1 bit per second, as shown in Figure 3. f is the frequency offset from the signal carrier frequency and Rs is the modulation symbol rate.

Note that the roll-off rate is only 20 dB per decade and is much slower than required by the NTIA mask, let alone the SFCG mask. Therefore, the require a relatively large guard band. When the 60 dB down guard band is included, the bandwidth efficiency becomes 0.00133 bits per second per Hz. One can improve the spectrum of the nominally square-shaped digital modulation time domain pulse by rounding its corners or by post-final HPA RF filtering or by both. These reduce the high-frequency components of the spectrum. A commonly used modulation pulse shaping results in an SRRC spectrum. This spectrum theoretically has no energy beyond +/-Rs(1+β)/2, where Rs is the symbol rate and β is the bandwidth expansion parameter of the SRRC signal. Figure 4 shows the normalized spectral density of SRRC M-ary PSK signals for a typical β value of 0.35. The upper curve is the spectrum of the SRRC only. It assumes that the data input to the SRRC filter is a series of positive or negative delta functions (infinite impulses). The lower curve is the output spectrum when one uses square pulses as the data input of the SRRC pulse filter. This cascade results in a spectrum that is the multiplication of the spectrum of the square pulses with the SRRC spectrum of the pulse shaping filter. A spectral floor of –60 dB is assumed.

6 National Telecommunications and Information Administration (NTIA), Manual of Regulations and Procedures for Federal Radio Frequency Management, rev. September 2006, Figure 5.6.1, pp. 5-44.

5 Figure 2. Space Services Permitted Unwanted Emission

0

-10

-20

-30

dBsd dBsd -40

-50 necessary bandwidth

-60 dB relative to the maximum value of power spectralpower density (psd) within the -70 10 100 1000 10000 Frequency offset as a Percent of the Necessary Bandwidth

Figure 3. Theoretical Spectral Density of Phase Shift Keyed Modulation with Square Pulses

0

-10

-20

-30

-40

-50

-60 Normalized Spectral Density (dB/Hz) -70 0.1 1 10 100 1000 Normalized Frequency Offset from Carrier

Since the modulation is applied with amplitude shaping to the quadrature inputs of the M-ary PSK modulator, the pulse shaping for spectral reduction will cause the amplitude of the signal to fluctuate proportionally with the modulation pulse during each symbol. This would not be a problem if the upconverters that set the output frequency of the signal and the HPA that sets the output power of the transmitter were perfectly linear. Linear means that the HPA output signal level is directly proportional to

6 the input level from the modulator over the desired amplitude range of the signal. However, a nonlinear HPA’s saturation partially flattens the modulation amplitude peaks and partially regenerates the frequency spectral side lobes. This process is called spectral regrowth.

When the spectral regrowth exceeds the required mask parameters or the limits reached by negotiations with adjacent band users, the nonlinearity and its spectral regrowth can be partially mitigated by several techniques. These include:

1. Using a pre-HPA linearizer that predistorts the signal to compensate as much as possible for the saturation curve that typically exists in power amplifers. 2. Reducing the output power at the operating point of the HPA relative to the saturated power output so that the linearity is improved at the cost of more output power back-off (OPBO). 3. Adding post-final HPA bandpass filtering, but this is at the cost of some loss in signal power due to the filter insertion loss.

Figure 4. Spectral Density of SRRC and SRRC*Sinc Pulse Shaping 0

-10 SRRC -20 Sinc*SRRC -30

-40

-50

Spectral Density (dB/Hz) Spectral -60

-70 0 0.20.40.60.81 1.21.41.61.82 Frequency/Symbol Rate

The regrowth of sideband spectrum has been examined using the NOAA-NESDIS GOES-R satellite series communications test bed to evaluate practical waveform performance in real hardware. We included an L-band, space-qualified design TWTA (purchased from Boeing) in the test bed to examine sideband regrowth. We implemented the modulator and demodulator in FPGA technology, including SRRC spectral filtering, and did some end-to-end bit error rate (BER) and spectral regrowth testing with and without linearization and variable amounts of OPBO. We examined these effects using a 16 APSK waveform. Figures 5 and 6 document the measured spectrum for these tests. Figure 5 shows the TWTA output spectrum with OPBO as a parameter. OPBO is referenced to the saturated power of the TWTA of 150 watts. The vertical axis should be treated as a relative power and not the absolute RF dBm power. No linearization was used in this case. Then the TWTA was mated with a commercial analog L-band linearizer at the input port and tuned for optimum sideband reduction. Figure 6 compares the data for the nonlinearized TWTA with that for a linearized TWTA. It shows that the main and first sideband spectrum levels for the linearized case with 4 dB OPBO matched that for the nonlinearized 7 dB OPBO case. The conclusion was that linearization in this case reduced the OPBO by 3 dB for the higher sideband levels. However, for the linearized case, the lower sideband levels are worse than those for the nonlinearized

7 case. Also the higher floor level is due to the high noise figure of the analog linearizer and is not caused by spectrum regrowth. This can be mitigated by use of a low-noise figure analog linearizer or better yet, by a purely digital linearizer. The peak spectrum levels have been set to the same level for easier comparison of sidebands.

Figure 5. Attenuated L-Band TWTA Output Power Spectral Density Versus Output Power Back-off

-15 2-dB OPBO 3-dB OPBO -25 4-dB OPBO

-35 5-dB OPBO 6-dB OPBO -45 7-dB OPBO 8-dB OPBO -55 9-dB OPBO

Amplitude, dB -65 10-dB OPBO

-75

-85

-95 1660 1670 1680 1690 1700 1710 1720 Frequency, MHz

Shannon’s limit dictates the maximum bandwidth efficiency at which one can transmit at a certain signal-to-noise power level with error-free communications. The continuous curve in Figure 7 shows this upper bound on the bandwidth efficiency in bits/second (bps) per Hz (capacity) versus power efficiency in energy per bit to noise density ratio (Eb/No). It also shows the efficiency for a selection of modulation and coding schemes for BER = 10–6.7 With modern modulation and coding techniques, one can communicate with a signal-to-noise level that is only about 1 dB greater than Shannon’s limit for a prescribed bandwidth efficiency. TC/16QAM (turbo code) in the graph (shown by the solid squares connected with a dotted line) approaches that value. Shannon’s limit also dictates that increasing the modulation alphabet increases the required signal-to-noise ratio. However, there is a large gap (on the order of 2 to 3 dB with coding) between the Shannon limit and the measured performance for most signals using readily available modulations and modest codes.

Adding error correction coding despite its small increase in bandwidth with high rate codes can mitigate that gap considerably in many cases. The expansion of the bandwidth can be minimized by the use of a high rate code. The code rate is defined as the ratio of the user data bit rate to the encoded bit rate. The newer more powerful turbo and low-density parity check (LDPC) codes in many cases can provide near–Shannon limit performance (within 1 dB) for code rates greater than 0.5 and up to 0.98. A minimum value suggested for GOES-R is a 0.876 code rate, which has been used in our GOES-R laboratory test bed. Commercially available hardware of code rate 0.876 and higher and consistent with the DVB-S series of standards has been available for years (refer to various commercial communications

7 Wang, Charles C., The Aerospace Corporation, personal correspondence.

8 satellite electronics equipment suppliers such as Comtech and Broadcom). Attainable bandwidth efficiencies are below the theoretical Shannon limit. Areas above the Shannon limit are not achievable. Processing delay and decoder chip complexity increase as one tries to communicate more closely to the theoretical limit.

Figure 6. TWTA-LTWTA Comparison - Normalized OPBO

-15 -20 TWTA 7-dB OPBO, -25 norm. -30 LTWTA 4-dB OPBO -35 -40 -45 -50 -55 -60

Amplitude, dB -65 -70 -75 -80 -85 -90 -95 1660 1670 1680 1690 1700 1710 1720 Frequency, MHz

Figure 7. Bandwidth Efficiency Versus Power Efficiency for BER = 10-6

10 TC: Turbo Code 8 CON_C: Concat. Code TC/QPSK CON_C/QPSK 6 Shannon TC/8PSK ) CON_C/8PSK Theoretical Uncoded TC/16QAM Limit 16QAM CON_C/16QAM 4 14.5 dB 13.9 dB Within each group of 4: 3 1 - (the leftmost one), Uncoded TC(R4/5;16K) or Uncoded 8PSK CC2/3-RS(255,223) QPSK 2 - (the 2nd from left), 3 4 TC(R8/9,16K) or 2 2 Data Rate/BW (bps/Hz Rate/BW Data 1 10.5 dB CC3/4-RS(255,223) 3 - (the 2nd from right), More Bandwidth Efficiency 4 TC(R11/12;16K) or A B 3 PC8/9-RS(255,223) 2 4 - (the rightmost one), 1 TC(R15/16,16K) or 1 PC8/9-RS(255,243) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Eb/N0 (dB) More Power Efficiency

9 NOAA requirements in increased sensor data rate from the current 2.66 megabits per second (Mbps) of the GOES-L and M series to roughly 150 Mbps as an upper bound for GOES-R through U series has necessitated a migration of the sensor data downlink from the legacy L-band usage up to X band, within the 8025 to 8400 MHz band. However, care must be exercised so that unwanted emissions from this signal do not impact ground reception of the adjacent 8400 to 8450 MHz band allocated to space research and particularly for deep space to Earth transmission. The GOES-R stage 2 NTIA filing with the NTIA stipulates 180 MHz of bandwidth. If QPSK without pulse shaping and rate 2/3 error correction coding were used, the necessary bandwidth would be 157 MHz. This would leave 23 MHz for the two- sided guard band for the filter to roll off the spectrum of the signal, or 11.5 MHz per side. However, with the bandwidth expansion factor, β = 0.35, for SRRC 16 APSK modulation and rate 0.876 coding, the necessary bandwidth is reduced by 46% to about 85 MHz plus the guard bands. This modulation and coding combination will allow a reduction of the complexity and signal loss associated with the HPA RF output filtering for diplexing the receiving and the transmission bands, but it will increase the required HPA output power after other adjustments by about 3 dB.

Spurious emissions are the other part of unwanted emissions. Out-of-band emissions are considered for the most part to be caused by the modulation process. The spurious emissions are considered to include intermodulation products, with such causes as nonlinear junctions, parasitics, and harmonics. Spurious emissions are usually controlled by filtering.

COMMUNICATIONS TECHNOLOGY STATUS

The advanced state of the art and ready availability of the above technologies and techniques are demonstrated in the DVB-S (1994), DVB-DSNG (1997), and the DVB-S2 evolution of Europe’s “Digital Video Broadcast (DVB)” series of standards and in the equipment of the world’s suppliers to its requirements. The DVB standard was developed for satellite television broadcast for high-definition TV and other applications. Standards are developed as well by the DVB body for cable transmissions. The 1997 version (DVB-DSNG) added 8 PSK and 16 QAM modulations. The S2 version is designed to squeeze high-definition video streams into existing 6 MHz channel bandwidths, including some guard band. “The DVB standards are maintained by the DVB Project, an industry consortium with more than 270 members, and they are published by a Joint Technical Committee (JTC) of the European Telecommunications Standards Institute (ETSI), European Committee for Electrotechnical Standardization (CENELEC), and European Broadcasting Union (EBU).”8

This S2 standard includes modulation modes up to and including 32 APSK. This means that there are 5 bits per symbol for the highest bandwidth-efficient mode. Other variants of APSK are 64 APSK, 128 APSK, and 256 APSK. LDPC coding is used with coding rates selected to complement the modulation modes for uniform steps in required Eb/No and bandwidth efficiency.

The DVB-S standard has been widely accepted and used in Europe, Asia, and the USA, with commercial equipment being designed and manufactured throughout the world. DVB-S2 provides a range of bandwidth efficiencies up to 4.5 bit/s/Hz. Comtech and other companies are pioneering the development of modems for ground applications that support DVB-S through DVB-S2 standard. See, for example, data sheets of their models SLM-5650, SLM-7650, SLM-8650, CLM-9600L, CDM-Qx, CDM- 600, CDM-700, and CDM-8000, all of which are available online.

In addition the European Space Agency (ESA) is developing a slightly different family of waveforms with the same family of modulation types as DVB-S2 but mates them with serial concatenated turbo coding (SCTC). Some variations include a Reed-Solomon (RS) or Bose-Chadhuri-Hocquenghem (BCH) outer code to reduce low BER Eb/No flaring.

8 Internet Wikipedia Encyclopedia for “DVB” May 25, 2007.

10

Further, the U.S. Department of Defense (DOD) issued a memorandum on February 10, 2006, establishing a requirement that DVB-S2 be used in DOD-owned and -leased satellite communications systems.9 In a paper presented at the 2006 annual Society for Photo-Optical Instrumentation Engineers (SPIE) meeting in San Diego, CA, the authors Cragg and Brockman discussed the decision of NOAA’s National Weather Service (NWS) to require the use of the DVB-S standard for any of its contracted commercial satellite data broadcasts to its field offices.10

In a 2005 e-mail communication with Roger Heymann of NOAA-NESDIS, Daniel Enns of Comtech reported (Daniel Enns is also an advisor on modulation and coding to the ATSC standards body in the USA):

At CEFD (Comtech EF Data) we have been providing 8PSK, 16QAM satellite modem solutions for over 10 years. We probably have well thousands [sic] satellite modems on the market today operating in 8PSK and 16QAM links. We have provided these satellite modems both for the commercial market as well as the Military/Government market. Our newer satellite modem i.e. [sic] the CDM 700 and the SDM 5650. We also offer 64QAM as an option for very high data rate throughput. We have also started to ship satellite modulators with the new DVB-S2 standard our CDM 710 [sic]. Comtech EF Data is also the leader in the market of Forward error correction, hence we (develop and market) modems that offer Viterbi, Viterbi+RS, Turbo Coding as well as LDPC/BCH. NOAA would certainly be well served by adopting higher order modulation and improved forward error correction (FEC) for the new satellite services.11

Several of the Department of Defense Air Force space programs are building or preparing to build satellites that will support on-board modulation, coding, demodulation, and decoding of 8 PSK and 16 APSK modulations as well as SCTCs. These are being prepared for flight by around 2014. They are currently building application-specific integrated circuits (ASIC) for these functions using technology that lends itself to transfer to space-qualified integrated circuit production lines.

FUTURE WORK

The residual technology development remaining for the modulation and coding for space applications such as the GOES-R series satellites is to complete the conversion of the above designs to space- qualified ASIC chips for the spaceborne transmitters and receivers through porting the ASIC design code from the commercial chip fabrication lines into the space-qualified lines. The hardware components for the ground-based applications already exist in commercial fabrication lines.

CONCLUSIONS ON CONTROL OF UNWANTED RF SPECTRUM EMISSIONS

Controlling unwanted RF spectrum emissions—meaning both out-of-band and spurious—is both a national and international problem with growing RF spectrum use and congestion. For U.S. federal bodies, the use of RF spectrum when crossing U.S. boundaries must meet both NTIA and ITU requirements. The NTIA and ITU both specify RF spectrum masks as guidelines to limiting RF power

9 Memorandum for Secretaries of the Military Departments, “Department of Defense Policy for Transmission of Internet Protocol DOD-Leased Lines and DOD-Owned Transponded Satellite Systems,” John Grimes, Chief Information Officer, DOD, Feb. 10, 2006 10 Published paper, “Evolution of the NOAA National Weather Service Satellite Broadcast Network (SBN) to Europe’s DVB-s Satellite Communications Technology Standard,” P. Cragg, NOAA NWS, et al., Published proceedings (manuscripts) of SPIE, “Satellite data Compression, Communications, and Archiving,” R. Heymann, C. C. Wang, Schmit Chairs/Editors, 13–14 Aug. 2006. 11 E-mail from Daniel Enns, Head of the ATSC, digital TV standards body for the U.S., to Roger Heymann, 8 February 2005.

11 emissions outside of a user’s assigned allocations. Further, in cases supplementing use of recommended masks by conducting negotiations with adjacent-band RF spectrum users seems logical. Spectrum efficiency plays a role in containing unwanted emissions. The use of advanced higher-order modulations as discussed in the paper requires less bandwidth and will decrease the required guard bandwidth to adjacent band users. However, if the strategy for requesting bandwidth is to base it on a lower bandwidth efficiency signal and then improve that efficiency after the allocation is obtained, then the available guard band would be increased and as such would reduce the out-of-band emissions on the adjacent users. Such advanced modulations are readily available in the commercial sector through equipment built to Europe’s DVB standards.

ACRONYM GLOSSARY

APSK amplitude phase shift keying ASIC application-specific integrated circuit ATSC Advanced Television Systems Committee BPSK binary phase shift keying BCH Bose-Chadhuri-Hocquenghem BER bit error rate bps bits per second CCSDS Consultative Committee for Space Data Systems CEFD Comtech EF Data CENELEC European Committee for Electrotechnical Standardization dB , defined as 10Log10 (power ratio) dBsd dB of spectral density relative to the peak spectral density DOD Department of Defense DSNG digital satellite news gathering DVB Digital Video Broadcast DVB-S DVB satellite DVB-S2 DVB-S second generation EBU European Broadcasting Union Eb/No energy per bit to noise density ratio EESS Earth exploration satellite services ESA European Space Agency ETSI European Telecommunications Standards Institute FCC Federal Communications Commission FEC forward error correction FPGA field-programmable gate array GOES Geostationary Operational Environmental Satellite GOES-L and M L and M satellites of GOES series GOES-R R satellite of GOES series GRB global rebroadcast HPA high-power amplifier ITU International Telecommunication Union JTC Joint Technical Committee LDPC low-density parity check LTWTA linearized TWTA Mbps megabits per second METSATS Meteorological Satellite Services MHz megahertz NESDIS National Environmental Satellite Service NOAA National Oceanic and Atmospheric Administration NTIA National Telecommunications and Information Administration NWS National Weather Service

12 OPBO output power back-off PDU processed data uplink PSK phase shift keying QAM quadrature amplitude modulation QPSK quadrature phase shift keying RF radio frequency RFI radio frequency interference RS Reed-Solomon SCTC serial concatenated turbo code SFCG Space Frequency Coordination Group SRRC square root raised cosine TWTA traveling wave tube amplifier UN United Nations U.S. United States, an alternate to USA USA United States of America

13