Common Data Link [CDL]
The Common Data Link (CDL) program is designed to achieve data link interoperability and provide seamless communications between multiple Intelligence, Surveillance, and Reconnaissance (ISR) collection systems operated by armed services and government agencies. CDL provides full-duplex, jam resistant, digital microwave communications between the ISR sensor, sensor platform, and surface terminals. The CDL Program establishes data link standards and specifications identifying compatibility and interoperability requirements between collection platforms and surface terminals across user organizations.
In 1979, the Common Data Link (CDL) Program foundation originated the Interoperable Data Link (IDL) program. The United States Air Force/Assistant Secretary of Defense (USAF/ASD) and the National Security Agency (NSA) developed the IDL the U-2 platform. In 1988, the Office of the ASD (OASD)/Command, Control, Communications, and Intelligence (C3I) recognized the success of the IDL program with a decision to develop a standard communication architecture that would be common across all Department of Defense (DoD) Services. OASD/C3I mandated the CDL proliferation to users involved in the collection and dissemination of wideband Intelligence, Surveillance, and Reconnaissance (ISR) data.
CDL is a full-duplex, jam resistant spread spectrum, point-to-point digital link. The uplink operates at 200kbps-and possibly up to 45Mbps. The downlink can operate at 10.71 -45 Mbps, 137 Mbps, or 234 Mbps In addition; rates of 548Mbps and 1096Mbps will be supported. The CDL family has five classes of links:
- Class I - Ground-based applications with airborne platforms operating at speeds Mach 2.3 at altitude up to 80,000 ft.
- Class II - Speeds up to Mach 5 and Altitudes up to 150,000 ft.
- Class III - Speeds up to Mach 5 and Altitudes up to 500,000 ft.
- Class IV - Terminals in satellites orbiting at 750nm.
- Class V - Terminals in relay satellites operating at greater altitudes
Current Class I Terminals (examples) include Land-based Modular Interoperable Surface Terminal (MIST) and the Sea-based Common Data Link-Navy (CDL-N) previously known as Common High Bandwidth Data Link Surface Terminal (CHBDL-ST). The latter supports the Advanced Tactical Airborne Reconnaissance System (ATARS) and the Battle Group Passive Horizon Extension System (BGPHES). CDL permits the remote operation and exploitation of sensors carried by CDL fitted platforms from BLOS locations via satellite. There are two satellite CDL systems in use Senior Span and Senior Spur. Senior Span makes use of a Span Airborne Data Link system operating in the I-band via Defense Satellite Communication System II/III spacecraft. Senior Spur meanwhile operates in the Ku-band in order to gain bandwidth. In August and November 1999, the USN demonstrated a phased array antenna that can handle both satcoms and CDL operation. Frequencies utilized were in the EF/S band (2.2-2.3 GHz) and I/X band (7.25-8.4 GHz) both for military satcom, plus I/X band (9.7-10.5 GHz) and Ku-band (14.5-15.35 GHz) for CDL and Ku-band (10.95-14.5 GHz) for commercial satcom. The USAF's UAV Global Hawk participated in the exercise Linked Seas 00 demonstrating downlinking of radar imagery to both the US Army's Tactical Exploitation System and to the USS George Washington, and subsequently to the Joint Analysis Center at Molesworth in the UKS. In the following Joint Task Force Exercise 00-02, CDL was utilized to pass retasking requests to Global Hawk from ship and land-based terminals. Previously the USN sponsored a demonstration of Synthetic Radar Imagery downlinking direct from the UAV via CDL into the Joint Services Imagery Processor Navy, passed through the Common Imagery Processor to the Precision Targeting Workstation (PWT) for analysis. Imagery was then provided to an airborne F/A-18.
The Tactical Common Data Link [TCDL] program provides a family of interoperable, secure, digital data links for use with both manned and unmanned airborne reconnaissance platforms. Possible platforms include UAVs, P-3 Orion, Guardrail, JSTARS, LAMPS, and etc. It will transmit and receive ISR data at rates from 1.544Mbps to at least 10.7 Mbps over ranges of 200 kilometers. TCDL will soon support the required higher CDL rates of 45, 137 and 274 Mbps.
Over the last decade, the CDL program has taken many successful steps towards the better, faster, and cheaper communication systems that will provide timely information to the Warfighter. Some of these successes include:
- High-density integrated circuit development such as Monolithic Microwave Integrated Circuits (MMIC), Field Programmable Gate Arrays (FPGA), Application Specific Integrated Circuits (ASIC), and Multi-Chip Modules (MCM) to lower cost as well as size, weight, and power.
- Migration from Mil-Standard to Commercial-Off-the Shelf (COTS) based equipment integrated into military communication systems such as the Tactical Interoperable Ground Data Link (TIGDL) and the Tactical Common Data Link (TCDL)
- High-Speed Data Network Interfaces and Wireless Data Link Networking development such as Asynchronous Transfer Mode (ATM), High-Speed Ethernet, and Fiber Distributed Data Interface (FDDI) networks.
- Numerous technology demonstration programs such as Predator Unmanned Aerial Vehicle (UAV), Global Hawk UAV, Real Time Sensor Data (RTSDL), Battlefield Awareness Data Dissemination (BADD) Demonstration, and Airborne Reconnaissance Low (ARL) Interoperability Demonstration.
- Satellite network connectivity development programs such as Sr. SPAN, Sr. SPUR, Contingency Airborne Reconnaissance System (CARS)/Mobile Stretch (MOBSTR), Predator UAV, Global Hawk UAV.
- Air-to-Air relay program/networking such as the Guardrail Multi-Role Data Link (MRDL) which includes the Direct Air to Satellite Relay (DASR)/ Repeater Program.
- Future architecture and waveform development such as the Airborne Information Transmission (ABIT) Program, which provides Low Probability of Intercept (LPI) and Low Probability of Detection (LPD) security features.
The basic CDL signal flow is from user interface to radio frequency (RF) output. The left side is a ground configuration and the right side is an airborne configuration. The basic CDL user interface is comprised of various modules (mux input channels) to a tunable RF output in the Ku band.
Multiplexer simultaneously provides multiple dedicated channels to users and mixes the data bits from each channel to form one aggregate bit stream. The method used to mix data from the channels is to organize the bits into multiplexer frames. These frames are then used to create the aggregate bit stream. A Multiplexer frame is a fixed length sequence of bits, with each user channel being allocated specific bit position within a frame. Synchronization (SYNC) code is a sequence of bits that is also included in each multiplexer frame. These SYNC codes provide the ability to identify each frame within the aggregate bit stream and the relative bit position within the frame. The Multiplexer provides a clock to each channel and accepts data bit according to the clock. The Multiplexer function is used in the Forward Link (FL) and in all modes of the Return Link (RL).
The Randomization function is used to ensure an even distribution of ones and zeros, which is necessary for optimal performance of data link functions as demodulation. Randomization occurs by exclusive Or Adding (XOR), which adds a bit from a specially selected bit sequence to each bit within the multiplexer frame, except for SYNC bits. The bit sequence that is used to randomize is referred to as pseudo-random or pseudo-noise (PN) sequence. The random distribution of the bit sequence matches a Gaussian distribution, which is a random sequence, which never repeats or ends. This function occurs simultaneously with the construction of each multiplexing frame. Randomization is performed in all modes of RL including NB, MB, and WB modes of operation, but is not done on the RL.
Encryption is used to protect the data in the event that the data is intercepted, and is performed on the aggregate bit stream by Communications Security (COMSEC) devices. COMSEC devices change (encrypt) the bit stream in such a way that is difficult to reconstruct the original bit stream without a complimentary decryption device. The disadvantage of encryption is that the decryption process adds multiple additional errors for any error encountered. This characteristic is known as error extension.
Most CDL hardware has the capability to employ COMSEC devices so that the data coming from the multiplexer can be encrypted. In order to lower cost, lower end CDL systems may not have this ability. Systems with encryption have been designed such that COMSEC devices can be physically bypassed if necessary. The KG-68 AND KG-135 family of COMSEC devices areas used as the CDL standard. CDL FL and Narrow Band (NB) mode of RL have traditionally been encrypted, while Medium Band (MB) and Wide Band (WB) modes of the RL have not been encrypted. With the development of the KG-135, a high rate encryption device, MB and WB modes of the RL may be encrypted in the future after some hardware modifications.
There are several types of encoders that are used within the CDL system, although differential and convolutional encoders are the two most common. Some types of encoders add redundant bits to the bit stream, which enable the correction of errors that may occur in transmission. The addition of redundant bits expands the data rate. Other encoders do not add redundant bits; rather they change the signal, which allows for the recovery of the original signal.
The purpose of differential encoding is to resolve signal (phase) ambiguities that are created through the demodulation process on the receive side of the data link. For reasons that cannot be explained within this document, the output of the demodulator creates a bit stream, which may be inverted from that which is transmitted. In either case of the demodulator, inverting or not inverting the bit stream, the transmission between ones and zeros are preserved without ambiguity. Differential encoding transforms the digital bit stream by converting space or zero pulses into the transitions that occur between a one and a zero. Since there is no ambiguity of the transmissions coming from the demodulator, the differential decoder can restore the bit stream. This process, however, doubles the bit error rate, because one bit error will affect two transitions, which correspond to a 3db loss to the system. CDL uses differential encoding on the FL segment and the WB modes of the RL segment. CDL does not use differential encoding in the NB mode of the RL.
Convolutional encoding is applied to the data link signal in order to correct bit errors that might occur during transmission and results in coding gain for the system. Decoding is the process that actually corrects transmission errors, but in order for decoding to occur, the signal must be properly prepared. Encoding prepares the signal for the decoder by adding selective redundant bits. The motivation for using encoding comes from the error extension characteristic associated with the encryption/decryption process. Through the convolutional encoding/decoding process, the majority of transmission errors will be corrected before they are passed onto the decryption process. Rate ½, constraint length seven characterizes the convolutional coding that is used within the CDL. Rate ½ is the comparison of the rate of bits entering to the rate of bits leaving the encoding process and in this case refers to the doubling of the data rate.
Reed Solomon Encoder
Traditionally, MB and WB modes of the RL segment have not been enclosed. In the future, Reed Solomon (RS) Encoding and the KGv-135 COMSEC will provide encoding and encrypting for MB and WB modes of the RL segment. As with convolutional encoding, RS coding adds redundant bits and creates code words that enable the decoding process to correct errors. RS differs from convolutional encoding by performing block encoding (using bytes) rather than the bit-wise encoding. Because of block encoding, RS is eight times faster in than convolutional encoding.
Interleaving is used to inter-mix the bits of the code words generated through convolutional encoding. The motivation for leaving interleaving is to compensate for burst or sequential errors, which can otherwise exceed the capability of the decoder to correct errors. Each code word generated through convolutional encoding can only correct a limited number of errors that occur in that code word. Sequential errors can cause multiple errors in a single code word, which can exceed the error correcting capability of the decoding process. Interleaving distributes bits so that if sequential errors do occur they will be distributed over multiple code words. For example, seven errors in a single code word will be distributed during interleaving into seven code words each having a single error. While the decoder may not be able to recover data in a code word with seven errors, it can easily recover a single error in seven code words.
The disadvantage of interleaving is the delay created by writing a block of bits into memory, inter-mixing the bits, and then pulling the bits from memory. This delay is dependent on the number of bits that are interleaved at a time and the data rate of the aggregate bit stream. Interleaving is performed only on a finite block of bits at a time. Similar to multiplexing, interleaving requires framing the aggregate bit stream and adding SYNC codes. Interleaving frames, SYNC codes, and the interleaving algorithm are all defined in the CDL Specification. Interleaving is currently used in the FL segment and the NB mode of the RL segment. Interleaving is not used in MB or WB modes of the RL.
Direct Sequence Spread Spectrum Modulator
Spreading is performed to provide jam resistance against narrow band jammers, which concentrate all power on the signal such as tone and narrow band noise. In addition, it effectively hides the signal spectrum, resulting in low probability of the detection/low probability of intercept (LPD/LPI) signal qualities. The aggregate bit stream is spread by exclusive or adding (XOR); combining the aggregate bit stream with a much faster pseudo-random bit sequence. To anyone who does not posses an exact synchronized replica of the PN code the signal appears to be noise. For various reasons having to do with vulnerability, and command and control priority, CDL spreads only the FL segment and does not spread any mode of the RL segment.
The modulator converts the aggregate digital bit stream into Radio Frequency (RF) analog signal. CDL uses binary Phase Shift Keying (BPSK) modulation for the FL and Offset Quadrature Phase Shift Keying (OQPSK) modulation for all three modes of the RL. CDL has traditionally modulated the digital bit stream into the Intermediate Frequency (IF) of 1700 MHZ which enables the same modulator to be used in all systems independent of the final transmit frequency. The modulation process can directly the bit stream to any other intermediate frequency or directly to the transmit frequency if desired. In some CDL systems a frequency of 300 MHZ is used as an IF.
The upconverter translates the 1700 MHZ RF signal to its final X or Ku band frequency where it is amplified to provide the required power to the antenna. Filtering assures that the spectral purity requirements for the allocated frequency are met. The size of the power amplifier is determined through Link Budget analysis and mission requirements. Upconversion, power amplification, and filtering exist in both the FL and all modes of the RL.
The diplexer contains filters which isolate the transmit frequency from the receiver frequency. This makes it possible for the transmitter and receiver to share a common antenna.
The antenna, with the transmitter amplifier, provides a significant portion of the gain required to close a link. Antenna and diplexer components exist in the FL and all modes of RL.
The diplexer makes it possible for the transmitter and receiver to share a common antenna. The size of the antenna determines how much RF power is captured and provides a significant portion of the gain required to close the link. Antenna and diplexer components exist in the FL and all modes of the RL.
The Down Converter translates the RF transmit frequency from the X OR Ku band to an IF. Downconversion, power amplification and filtering exist both in FL and all modes of the RL.
The demodulator converts an RF analog signal into an aggregate digital bit stream. CDL use BPSK to demodulate the FL RF signal and uses OQPSK to demodulate all three RL modes. CDL has traditionally used an IF in order to use the same demodulator independent of the received RF frequency. The demodulator process generates phase ambiguity when generating the aggregate digital bit stream from the analog RF Signal. The ambiguity is that the system does not know whether the two signals coming out of the demodulator (I and Q) are inverted or are swapped. The ambiguity is resolved in the digital processing section of the data link, either through differential decoding or some other means. When convolutional encoding is used, the demodulator produces three digital bits representing the analog signal, also known as three bit soft decision. The Viterbi decoder resolves the three soft decision bits into the one hard decision bit. Allowing the decoder to resolve the soft decision provides additional coding gain to the data link. Demodulation is a required function in the FL and in all modes of the RL.
Direct Sequence Spread Spectrum Modulator
De-spreading is the process that extracts the aggregate bit stream from the much faster PN signal bit stream. The process mixes the aggregate bit stream with a much faster PN code, which hides the power spectrum within the noise of the frequency band. Once the signal is spread, any narrow band jammers that are encountered during transmission will be mollified through the de-spreading process. De-spreading occurs by finding correlation with the known PN code. To anyone who does not posses an exact synchronized replica of the PN code the signal appears to be noise. The de-spreading process occurs only on the receive side of the FL.
For any bit stream that has been interleaved before transmission, the process, of de-interleaving is required to re-assemble the code words created by the convolutional encoder. Any sequential errors that may have occurred prior to de-interleaving will be distributed in pseudo-random fashion. The actual distribution characteristic of the errors is dependent on the algorithm used to interleave and de-interleave. The algorithm used within the CDL is designated to approximate that of pseudo-random distribution.
De-interleaving occurs by acquiring the SYNC codes that are placed into the interleaving frames by the interleaver. Once SYNC has been located, interleaving frames are identified and the de-interleaving process is performed on the block of bits contained in the frame. Generally, all three soft decision bits coming from the demodulator are de-interleaved simultaneously. Interleaving is currently performed on the FL and the NB mode of the RL. Interleaving is not used in MB or WB modes of the RL.
Decoders within the receive side of the data link must match specifically to the encoders used within the transmit side of the data link. There are two different types of encoding/decoding techniques that are used within the CDL systems. One type of encoding adds redundant bits, not only expanding the data rate, but also providing the information necessary to recover the error that may occur through transmission. The second type of encoding does not add redundant bits but transforms the waveform, which when reconstructed provides essential information that would be lost otherwise.
Viterbi decoding is applied to the data link signal to correct errors that may have been encountered through RF transmission. The encoding/decoding process adds what is referred to as coding gain, which may be necessary for the successful data link transmission. Decoding corrects errors before bypassing the aggregate bit stream on to the COMSEC decryption device. Correction of the errors occurs because the convolutional encoder (or transmit side of the data link) creates code words, which contain data bits with added redundant bits. The redundant bits allow the decoder to detect and correct errors that may exist in each code word. The process of decoding is much more complicated than the encoding process, and limits the speed of the bit stream that can be decoded. Since the FL and the NB mode of the RL are convolutionally encoded, both are required to be Viterbi decoded. MB and WB modes of the RL segment do not use Viterbi decoding.
If the aggregate bit stream has been differentially encoded on the transmit side of the data link, differential decoding will resolve the phase ambiguities created through the demodulation process. The differential encoding process does not introduce redundant bits, but transforms the waveform by converting the space signal (zeros) into transitions. Accordingly, the decoding process converts transitions back to spaces. Since a single bit error affects two transitions, differential decoding process doubles any bit error, corresponding to a 3dB loss to the system. Differential decoding is used on the FL and the MB and WB modes of the RL, but not used in the NB mode of RL.
Decryption is the process that reconstructs the original signal, which was altered through the COMSEC encryption device in the transmitter. Encryption/decryption provides protection for the signal in the event of interception. However, this protection is not free because of the error extension characteristic of COMSEC devices. Error extension occurs within the decryption process, which adds multiple errors for every bit error encountered. Decryption is a required process on the receive side if the bit stream was encrypted on the transmit side of the data link. Since CDL encrypts the FDL an NB mode of RL, these must be decrypted in the receiver. Bypassing decryption is allowable only when encryption is bypassed on the transmit side of the data link. MB and WB modes of the RL segment are not encrypted so decryption does not currently occur within MB and WB modes of the RL segment. With the development of the KGV-135 high rate encryption device, MB and WB modes may be encrypted in the future, requiring decryption on the receive side of the data link.
When randomization is performed on the transmit side of the data link, the aggregate bit stream must be de-randomized on the receive side. The de-randomization function is performed in all modes of the RL, but is not used within the FL. De-randomization is performed simultaneously with the de-multiplexing function. The first step of the de-randomization process is to acquire the multiplexing frame SYNC codes. Since SYNC bits are not randomized, SYNC acquisition can occur before de-randomization. Sync bits identify the multiplexing frame, which is de-randomized by modula-2 adding the same PN sequence that was used to randomize the frame to the bits within the frame.
De-multiplexing is the receiver function that re-creates the user channels from the aggregate bit stream. In order for the de-multiplexer function to operate properly, the aggregate bit stream must be organized into multiplexing frames. A multiplexing frame is a fixed length sequence of bits, with each user channel being allocated specific bit positions within the frame. Also included in each multiplexer frame are bits used as synchronization (SYNC) codes that identify the frame and provide the relative bit positions within the frame. The de-multiplexer scans the aggregate bit stream for SYNC codes. Since SYNC codes are placed in the same position of each frame, once the SYNC code in the first frame is located, the de-multiplexer can easily identify the following SYNC codes and the frames that contain them. Once synchronization has occurred, the de-multiplexer distributes the bits within the frame bit positions to the appropriate user channel. The de-multiplexer provides both the bit stream and then associated CDL clock to the User egress panel. The de-multiplexing function is used on the FL and all modes of RL.
|Join the GlobalSecurity.org mailing list|