Network Centric

NCW Warfare
101NETWORKED
Fundamentals Part
3

OPERATIONS Dr Carlo Kopp

Contrary to commonly held belief, JTIDS/MIDS is There are two basic categories of spread spectrum
JTIDS / MIDS not new technology, even if many current
production terminal equipments are recently
techniques, and both are used in JTIDS/MIDS.
Frequency hopping spread spectrum techniques
designed. The JTIDS modulations and protocols were the first to be introduced and the most widely
The Joint Tactical Information Distribution were devised during the 1970s in the US and NATO used in military communications.
countries, using a combination of modulation The basic idea underpinning all frequency hopping
System (JTIDS) network, and its evolved
technique invented in 1942, error correction coding radios is that the frequency or wavelength of the
offspring MIDS (Multifunction Information invented in 1960, and a timesharing technique of radio carrier wave continuously hops around over
Distribution System), form the defacto similar vintage. time. Typically, a pseudo-random coding scheme is
standard military digital network in use JTIDS/MIDS is also a limited networking scheme in used to determine the next frequency to which the
today – dominating military network terms of coverage footprint and achievable network carrier wave should hop. Unless a receiver knows
capacity or throughput. It’s architecture initially where the next hop will be, it cannot capture the
installations in the coming decade until aimed to provide situational awareness data and signal and decode it. A hostile intercept receiver
its planned replacement, the Joint targeting data in air defence operations, and this sees a carrier wave popping up and disappearing
Tactical Radio System (JTRS), finishes has constrained its utility in a number of ways. continuously over time, within some range of
development and is deployed in These limitations aside, JTIDS/MIDS provides frequencies unique to the radio design.
valuable capabilities, many of which have never Frequency hopping is used since it is very effective
sufficient numbers to matter. been seen before. Perhaps the most important of at frustrating hostile radio jammers. In a
these include transparency, the ability to network conventional radio scheme, the carrier wave sits
assets without significant operator intervention, constantly at some operating frequency and a
and ubiquity (the ability to connect air, land and sea hostile jammer can be easily tuned in to interfere
assets seamlessly). with it. As the pseudo-random hopping code is kept
To gain a good picture of the strengths and secret, only authorised receivers knowing that code
limitations of JTIDS/MIDS, we need to explore three can anticipate where the frequency hopper will hop
aspects of the design. The first is how it encodes to next. Without this knowledge the jammer is
and protects data, the second how it addresses frustrated.
FREQUENCY individual network terminals, and the third the Because radio waves travel at the finite speed of
geometrical constraints on its coverage. Inevitably light, a frequency hopper that hops quickly creates
many of its design features overlap these three much difficulty for a hostile jammer. By the time the
aspects, which has historically been a cause of hopped signal has propagated from its source to
much confusion in the uninitiated. the jammer, it is apt to have hopped again to a
different operating frequency. The jammer is thus
forced to play a futile game of catch up.
How JTIDS/MIDS There is a price to pay for the increased jam
carries data resistance of the frequency hopper. Because the
frequency hopper only uses a small fraction of the
available radio bandwidth at any time, the amount
JTIDS/MIDS is what is termed a 'Spread Spectrum of data it can carry is reduced accordingly. A rough
FREQUENCY HOPPER
TIME Multiple Access' system, as it uses spread measure is that a tenfold improvement in jam
spectrum radio techniques and provides a resistance is paid for by a tenfold reduction in data
FREQUENCY
mechanism for multiple terminal access. This throughput.
simple language conceals considerable complexity. Frequency hoppers are not immune to jamming,
Spread spectrum modulation techniques were but force up the cost of jamming equipment
invented by Hollywood actress Hedy Lamarr and considerably since the jammer has to emit a
musician George Antheil in 1942 (U.S patent jamming signal in each or most of the frequency
2,292,387) who discovered the technique while slots the frequency hopper jumps between. In
using a player piano to control frequency hops, but effect, an opponent has to maintain a battery of
did not achieve prominence until the middle of the jammers to straddle much of the frequency range
Cold War, as the complexity of such designs made the frequency hopper operates in. Ten times as
them expensive to build. much jam power means a ten times bigger
jamming system.
TIME
HYBRID FH/DS SPREAD SPECTRUM SYSTEM
(JOINT TACTICAL INFORMATION DISTRIBUTION SYSTEM)

12 DefenceTODAY magazine
 The second spread spectrum technique used is
termed 'direct spreading'. It is simpler than

NCW
frequency hopping and is also widely used. In a
direct spreading design each digital '1' or '0'
transmitted is replaced by a pseudo-random string
of '1's or '0's. Unless a receiver knows what the
pseudo-random encoded string is apriori, it has no
means of knowing whether the data sent is a digital

101
'1' or '0'.
Like frequency hoppers, direct spreading systems
can have good resistance to jamming. The rough
measure is that jam resistance is improved by a
ratio equal to the number of '1' or '0' transitions in
the pseudo-random sequence used to encode the
direct spreading message. As with frequency
NETWORKED hoppers, the price to be paid is a reduction of
achievable data throughput per radio bandwidth, in
OPERATIONS proportion to the length of the pseudo-random
spreading code.
As an example, a conventional radio using 100 MHz
of radio bandwidth might carry 150 Megabits/sec JTIDS Frequencies. JTIDS hops between a total
of data, but can be easily jammed by a hostile of 51 seperate frequencies (US Navy).
signal of similar strength. A spread spectrum radio,
which uses the same 100 MHz of radio spectrum,
might only carry 1 Megabit/sec of data, but can Without knowing both of these pseudo-random
cope with nearly 100 times more hostile jamming spreading codes an opponent can neither intercept
power before it gets into difficulty. the signal nor retune a jammer quickly enough to
In summary, spread spectrum techniques can jam efficiently.
provide vastly better jam resistance than If we look at the baseline JTIDS/MIDS system in
conventional digital radio links, but per given radio perspective, it is transmitting tiny 5-bit chunks of
bandwidth pay for this in a proportionate reduction data 77,000 times per second, these chunks each
in how much data they can carry. Spread spectrum encoded pseudo-randomly and hopped between
radios can be intercepted only if the opponent 51 different radio frequencies pseudo-randomly.
knows what pseudo-random spreading codes are It is this mechanism that provides JTIDS/MIDS
being used. networks with good jam resistance and reasonably
Spread spectrum techniques good security.
Bit Duration Chip Duration have another interesting
feature also exploited in
Data Message JTIDS/MIDS. This feature is How JTIDS/MIDS
contingent on the
mathematical properties of the
addresses stations
PN Code pseudo-random codes being
used. If these codes have a The JTIDS/MIDS system provides a shared channel
property called 'orthogonality', between numerous stations. To allow stations to
Encoded Message
where a mathematical share the channel and be able to uniquely address
DIRECT SPREADING (DS)
operation called 'correlation' each other, another mechanism is required. This
between any two codes mechanism is termed Time Division Multiple
produces a result of zero, then Access (TDMA), and has been widely used in
two or more spread spectrum radios can operate commercial digital communications since the
DIRECT SPREADING
(JTIDS CCSK) AMPLITUDE within the same bandwidth at the same time. Each 1960s.
radio sees its peer's signals as little more than TDMA schemes rely on the idea of dividing time
background noise. into a large number of typically identical time slots.
Again there is a price to be paid. This is because Each specific channel to be carried is then
the jam resistance is reduced in proportion to the allocated its own time slot. When that time slot
number of spread spectrum radios with unique comes up, a transmission is sent and received by
FREQUENCY HOPPING FREQUENCY
(JTIDS FH) codes sharing the same radio bandwidth. As the pair of stations sharing the slot. All stations
AMPLITUDE
always there are no free lunches in this game. each take their turn, according to the
The baseline JTIDS/MIDS system hops at around preprogrammed slot allocation. The rest of the time
77,000 times per second. Each hop puts it into one both stations do nothing, waiting for their slot to
of 51 radio frequency slots, each separated by 3 arrive. These schemes are inherently 'cyclic', in
MegaHertz. The slots are fixed in the L-band, that, the sequence of slot allocations repeats again
FREQUENCY
shared with IFF secondary radar signals, starting at and again. If the rate at which these repetitions
SPECTRAL BEHAVIOUR 969 MegaHertz and ending at 1206 MegaHertz. occurs is fast enough, a user communicating
Two blind 'notches' are excluded to allow IFF to through a channel using time slots in this system
share the radio bandwidth. simply sees a channel that can carry however
Within each hop of the baseline JTIDS/MIDS many bits per second of data.
system, the signal is further encoded by way of All TDMA schemes must have a protocol that
direct spreading techniques, using a specific defines when time slots start and stop, and who
method termed Cyclic Code Shift Keying (CCSK). can use which time slot. Without such a protocol,
This second layer of 'spreading' converts 5 bits of chaos would be inevitable.
raw digital data into a 32-bit pseudo-random In the baseline JTIDS/MIDS system a twelve
sequence, transmitted in a short 6.4 microsecond second cycle is used, divided into 1536 individual
'pulse' (effectively a tiny burst transmission). time slots. Each time slot thus has 7.8125-
millisecond duration.

14 DefenceTODAY magazine
 Within each of these time slots a JTIDS/MIDS Up to this point we have simply explored the
station transmits multiple pulses each of 5-bits of mechanisms JTIDS/MIDS uses to create a

NCW
data. The standard specifies either 72, 258 or 444 transparent digital pipe between stations. What is
pulses per time slot. Each of the 6.4 microsecond being sent through that pipe adds further
duration pulses is separated in time by at most 6.6 functionality, and complexity, to the system.
microseconds.
What this means in practical terms is that each slot
permits the transmission of a chunk of raw data of The Link-16/TADIL-J

101
between 360 and 2220 bits in size every 7.8125
milliseconds, for a raw data transfer rate between
messaging format
46.08 kilobits/sec and 284,160 kilobits/sec. That
data rate is shared between all of the stations Like most modern digital communications
participating in the network. protocols, the Link-16 protocol transmits data in
Why the gaps between pulses, and why the sedate discrete and tightly defined chunks, termed
cyclic rate for the times slots. This reflects several messages. Each of these messages contains a
NETWORKED realities. The first is that the JTIDS design was
conceived to support interceptors orbiting over
data 'payload' and additional bits to facilitate its
use. While the data payload is the useful content,
OPERATIONS Germany intended to stop waves of Soviet fighters. the system cannot function without the other
Updating each interceptor with threat data once components of the message.
every twelve seconds is a reasonable worst-case The JTIDS system has seen ongoing evolution of its
number. The gaps between pulses were set to message formats. Early JTIDS Class I terminals
reflect 1970s receiver designs, which require finite fitted to E-3A AWACS and F-15C fighters used a
time to hop frequencies, and to minimise the time message format called Interim JTIDS Message
available to jammers. If you are hopping Format (IJMS), which has been superceded in later
frequencies every 13 microseconds and Class II terminals with the full Link-16 message
transmitting for only 6.4 microseconds, a jammer format defined in the US Mil-Std-6016 standard.
must be at a distance of much less than 2 km if it TADIL-J is a US Navy designation, Link-16 US Air
is to have any hope of jamming the tail of each Force and NATO. Typically, only later US Air Force
pulse, having detected the beginning of the pulse. terminals are compatible with both the IJMS and
JTIDS/MIDS however uses a third mechanism to Link-16 message formats.
cause grief to opponents, although it is not always Link-16 messages come in a variety of formats.
used. If we assume a pulse every 13 The essential tradeoff is the data throughput
microseconds, and no more than 444 pulses per versus jam resistance. Message packing formats
each 7.8125 millisecond time slot, we find a visible that carry less data have better jam resistance.
discrepancy in the numbers. Only 5.8 milliseconds All Link-16 messages share some common
are being used for actual transmission (3.354 for features, an essential byproduct of the need to
258 pulses and 0.9 for 72 pulses), against a slot synchronise receivers in terminals and to uniquely
duration of 7.8125 milliseconds. address terminals.
The remaining time is there to provide a time The basic structure of all Link-16 messages is that
allowance for the radio signal to propagate to all of a block of 36 synchronisation and header double
stations in the network, and to allow intentional pulses followed by the actual message payload.
random jittering of the time at which the The 16 double pulses allocated to the
(US Navy Image)
transmission starts in order to further confuse synchronisation function allow a receiver to lock on
eavesdroppers. to the JTIDS transmission prior to demodulating
The minimum time available for propagation and and decoding the transmission. The 16 double
jittering is 2 milliseconds, during which the signal pulses comprising the header contain information
JTIDS WAVEFORM
can travel 315 nautical miles less jitter time. In that identifies the message. An additional four
1536 SLOTS practice, a JTIDS/MIDS footprint of around 300 NMI double pulses are included to allow control of
7.8125 MSEC DURATION EACH
is assumed. timing.
The JTIDS/MIDS system is a good example of a The actual payload then contains either digital data
layered approach to resisting hostile jammers, for transmission between computers using the
0.936 MSEC 2.418 MSEC 4.4585 MSEC LESS JITTER carefully melded with a relatively conventional link, digitally encoded voice communications, or a
TDMA scheme for sharing bandwidth. unique message for Round Trip Timing (RTT). The
16 DP

PROPAGATION GUARD
16 DP
SYNC

HDR
JITTER

MESSAGE DATA
(VARIABLE)
Because of the type of pseudo-random codes system can also add redundant data to protect the
LINK−16 MESSAGE
TIME REFINEMENT 4 DOUBLE PULSES
used, the JTIDS/MIDS system allows 128 unique message from bursts of transmission errors,
SYNCHRONISATION PREAMBLE
VARIABLE JITTER PERIOD hopping codes. If more than one of these codes is typically as a result of hostile jamming.
DATA in use at the same time, the network is said to be This Error Detection And Correction (EDAC)
16 DP
16 DP

PROPAGATION GUARD
SYNC

HDR
JITTER

93 DOUBLE
(VARIABLE)
PULSES
'stacked'; it is concurrently operating between two mechanism uses Reed-Solomon (R-S) 15/31
Standard Double Pulse (STDP)
sets of network terminals and sharing the encoding which provides the ability to correct up to
bandwidth between them. The technique of 50 per cent of the encoded data if it is corrupted in
DATA DATA
'stacking nets' is widely used but can often result transit. This is achieved at the cost of committing
16 DP
16 DP

PROPAGATION GUARD
SYNC

HDR
JITTER

93 SINGLE 93 SINGLE
PULSES PULSES (VARIABLE)
in reduced throughput since the degradation in the 31 bits of message to carry only 15 bits of actual
Packed−2 Single Pulse (P2SP)
effective noise floor (see above) can result in content. Yet again jam resistance is improved, but
increased data transmission error rates. at the cost of halving throughput. Voice channels
DATA DATA
When 'stacking nets' the cited limit is usually 20 do not use R-S EDAC capability. It is worth
16 DP
16 DP
SYNC

HDR

93 DOUBLE 93 DOUBLE PROPAGATION GUARD

PULSES PULSES

Packed−2 Double Pulse (P2DP)

concurrent hopping patterns before error rates observing that Reed-Solomon coding is the basic
produce significant impact. In the presence of technology used in CD and DVD data protection.
jamming this will inevitably be reduced. If we The payload is made up of chunks of data termed
DATA DATA DATA DATA
16 DP
16 DP

assume each MIDS/JTIDS net is carrying data at Link-16 words, each of which contains 70 bits of
SYNC

HDR

93 SINGLE 93 SINGLE 93 SINGLE 93 SINGLE PROPAGATION GUARD

PULSES PULSES PULSES PULSES

Packed−4 Single Pulse (P4SP) 54 kilobits/sec (STDP see below), then stacking 20 data and 5 parity bits for protection. All Link 16
x 54 kilobits/sec yields and aggregate capacity of messages are made up of integer multiples of
about 1 Megabits/sec. For comparison, a single Link-16 words.
LET High Data Rate Waveform
16 DP
SYNC

HDR

40, 61, 93, 108, 123 Link−16 Words

PROPAGATION GUARD
digital TV channel with MPEG encoding consumes There are four basic Link-16 message formats
Link−16 Enhanced Throughput (LET)
(c) 2005, Carlo Kopp

twice that capacity. used. Some of these transmit every data pulse

16 DefenceTODAY magazine
 twice to achieve 100 per cent redundancy for
improved jam resistance, but also to compensate JTIDS access methods

NCW
for propagation problems (for technical readers:
CCSK is considered susceptible to multi-path While MIDS/JTIDS is a time division multiplex
interference) or antenna coverage limitations in system, in which timeslots are allocated to
manoeuvring platforms. individual users for data or voice transfers, the
The Standard Double Pulse (STDP) message format system requires further enhancement to permit
has the lowest throughput but best jam resistance. more flexibility in an environment where many user

101
It is typically used to carry three or six 70-bit Link- terminals may need access. In comparison,
16 words, permitting each slot to carry 210 or 420 commercial TDM systems typically lack this
bits of data. capability and are designed with quite rigid
The Packed-2 Single Pulse (P2SP) message format schemes for allocation of slots.
doubles throughput compared with the STDP MIDS/JTIDS timeslots are typically reserved as
format, but does so at the loss of jam resistance Transmit, Receive or Relay Transmit slots. A
provided by redundant double pulse transmission.
NETWORKED It carries six 70-bit Link-16 words per slot.
terminal cannot transmit in a slot reserved as a
Receive Slot.
OPERATIONS The Packed-2 Double Pulse (P2DP) message
format doubles throughput compared with the
There are four most commonly used Access
Methods in typical MIDS/JTIDS networks.
STDP format, but sacrifices the jitter capability, The Dedicated access method allows only a
again at the expense of jam resistance. It also specific terminal to transmit in a designated slot
carries six 70-bit Link-16 words per slot. and all others are only allowed to receive in this
The best throughput is provided by the Packed-4 slot. A tanker aircraft broadcasting its location and
Single Pulse (P4SP) message format, which has the fuel state might use the Dedicated Access Mode,
weakest jam resistance, as the double pulse while fighters would listen in this slot to monitor the
redundancy and jitter are not used. tanker.
These message formats determine how much data The Dedicated With Time Slot Re-use access
is sent in each time slot. They do not define what method is similar to the Dedicated Access Mode,
data is sent; that is defined by the message type. but allows a commander to reallocate the specific
slot to a particular terminal.
Link-16 message types The Contention Access Method mode allows all
terminals to transmit in a so designated slot. If a
'collision' occurs when two terminals try to access
Link-16 is characteristic of modern military the slot simultaneously, then the terminal with the
datalink designs, in that, it uses many dedicated more powerful signal wins.
message types for specific purposes, in addition to The Push-To-Talk (PTT) access method is used for
voice channel capabilities. voice channels. Stations will not access the slot
The Precise Participant Location Identification until it is vacant and no other terminal is
(PPLI) message type is widely used and a good transmitting in it.
example. This message type contains mission unit
identification (JTIDS Unit (JU), IFF codes, unit type,
mission identifiers, platform location and platform Limitations and
velocities, navigation accuracy, and datalink status.
The Round Trip Timing (RTT) message type is used
strengths of
to maintain timing synchronisation between JTIDS JTIDS/MIDS/Link-16
terminals in a network. RTT-I interrogation
messages are usually generated by platforms that
have difficulty synchronising; an RTT-R response JTIDS/MIDS/Link-16 provides tremendous
message is then sent by a platform with more flexibility and has proven to be very useful
accurate timing to enable synchronisation to be operationally. The US now uses it largely as a
corrected. substitute for classical IFF in combat. A number of
Each JTIDS terminal can support two digitised weapon datalinks recently trialled have been built
voice channels, termed "Voice Group A" and "Voice as derivatives of JTIDS, using unique message
Group B". Typically these channels are stacked on formats but retaining the modulation scheme.
different nets. This message type uses 930 bits in Of all of the established datalinks,
each slot for digitised voice. JTIDS/MIDS/Link-16 is usually regarded to be the
The MIDS/JTIDS system most jam resistant. The combination of frequency
uses encryption hopping, CCSK direct spreading, message random
techniques to protect jittering, double pulse redundancy and Reed-
payloads. Crypto variables Solomon coding essentially defeats simpler
(numbers) are used to jamming techniques and requires a combination of
select the pseudo-random high jam power, large jamming bandwidth and
frequency-hopping smart modulation to produce serious jamming
pattern, the jitter time, and effect.
the spreading pattern for JTIDS/MIDS/Link-16 is not however without its
each pulse. Additional limitations and problems. The limited data
crypto variables are used throughput is inherent in its basic architecture,
to encode the message which makes it ill suited to the transmission of bulk
payloads. Terminals have data such as ISR imagery or live video feeds. This
the architecture to permit constrains its usage to situational awareness
cryptographic separation functions, command and control, low data rate ISR
between nets being used functions, and derivative functions such as weapon
concurrently. guidance. In the long term it is likely to remain used
for these purposes.
(US Navy Image)

18 DefenceTODAY magazine
A second key limitation is that it is not well adapted to carrying ‘bursty’
computer traffic, as its ability to rapidly reallocate slots is limited. This is The future
another inherent limitation of TDM systems and not one easily changed.
The third key limitation is its complexity, which drives up demands for skill The long term US plan is to replace MIDS/JTIDS terminals over time with
levels in managing networks. While a user of a terminal might get by with the new Joint Tactical Radio System (JTRS or 'jitters'), which will exploit
training, which covers modes and message formats, personnel who are newer technologies such as ad hoc networking. JTRS remains in
required to configure and manage networks require extensive and deep development and delays have affected early production of the basic
training to be proficient – especially in environments where jamming is equipment.
expected. In the mean time, some enhancements are being introduced to
Another issue frequently identified as a problem is a propensity for overly MIDS/JTIDS to alleviate - but not solve - problems arising from its basic
enthusiastic use of net stacking resulting in transmission errors and design.
unreliability. This in part relates to the previous limitation and the inherent The Timeslot Reallocation (TSR) method is an additional Access Method
issues when performing multiplexing of spread spectrum channels. Every (see above) which attempts to automate reallocation of available timeslots
time an additional net is stacked, some jam resistance is lost, as the noise depending on immediate demand. What TSR does, via software
floor seen by platforms operating at the edge of the network pushes them enhancements in terminals, is to allow terminals to advertise their
closer to viable operating limits. immediate demand for network capacity to all other terminals, upon which
While JTIDS/MIDS/Link-16 is often seen as a panacea, it is not, and using an algorithm in software in used to determine how many each terminal
it successfully requires considerable insight into its idiosyncrasies can actually get.
Another enhancement is the Link-16 Enhanced Throughput (LET)
capability developed by Viasat in the US – intended to increase throughput.
An LET capable terminal can communicate with non-LET capable
JTIDS/MIDS/Link-16 terminals terminals, but not vice versa in LET mode.
LET works by replacing the spread spectrum and Reed-Solomon encoding
A detailed survey of JTIDS/MIDS/Link-16 terminal equipment is a theme in with a newer Reed-Solomon/Convolutional coding scheme, which can
its own right. adapt to required link capacity. LET can provide 3.33, 5.08, 7.75, 9.0 and
Early JTIDS terminals were prohibitively expensive, limited to IJMS format 10.25 times more throughput than the basic JTIDS modulation, but it does
messages, often bulky and used only for key platforms. The more recent so at the expense of both jam resistance and transmission range. The
high volume production MIDS Low Volume Terminal (MIDS-LVT) is much fastest LET mode may be unusable in many combat environments.
cheaper and more compact. Finally, encroachment into the JTIDS/MIDS portion of the L-band spectrum
Designed for use across a wide range of platforms, especially aircraft, a by civil operators will force the introduction of a 'frequency remapping'
typical MIDS-LVT terminal is designed as a 'swap-out' form factor capability in future terminals, where the 51 hop frequencies are remapped
replacement for existing TACAN terminals, retaining an embedded TACAN to avoid the frequencies used by civil operators.
transceiver as an option to cut integration costs, as the MIDS-LVT terminal
can reuse existing power, cooling, antennas and cabling. It requires a Mil-

NCW
Std-1553B or other connection to the central mission computer, and local
machine software to access the network. Typical hardware costs for this
class of terminal are around $250,000 per terminal, with around $250,000
per platform in software code to access the terminal. Additional software
integration costs may arise. Putting MIDS-LVT terminals into 35 aircraft
thus costs of the order of $20 million, or about $600,000 per aircraft.
Some estimates cited in Australia for network integration have been
Part 4 next issue ..
101
NETWORKED
ridiculously high, suggesting that the actual costs of JTIDS/MIDS/Link-16 OPERATIONS
terminal integration are not widely understood.

DefenceTODAY magazine
19