20111 Stevens Creek Blvd.

, #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Lanwave Technology, Inc.

Code Division/Spread Spectrum

“CDMA for consumer electronics”

Advantages over DSS, DECT (FH-TDMA) and W-CDMA

Code Division Spread Spectrum is one form of CDMA communication technique optimized for the cost
conscious, high volume and short cycle consumer electronics market. Its primary application is in digital
cordless telephony today. Meanwhile, it is being applied to wireless devices for the last 100-meter
connection to the Internet, and in the wireless local loop to provide sub-$20 handset cost, tariff free mobile
voice services in some third world countries.

Why is CD/SS technology suitable for consumer electronics? And how does it compare to other long-range
wireless transmission techniques such as DSS, DECT and Wide band CDMA? The next 10 slides will
introduce the history, theory and practice, and highlight some technical areas for comparison.

Consumer electronics are, by heritage, an “Open Architecture” market. All forms of technology finds its
niche, which eventually help every other technology. This philosophy drives the Lanwave view of this
market, and forms the cornerstone to our business approach from sales to research and engineering.

Lanwave © 2001 Page 1

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

The primary
difference between
analog and digital
Wireless Communication transmission is in the
addition of a coding
engine known as the
wireless DSP base
Input signal
(analog in Hertz)
Received signal
(analog output)
band chip. In the
Analog cordless phone and
Up conversion Down conversion
Transmission
(multiply by a carrier frequency) Channel (multiply by the same carrier frequency,
filtering and recovery.) similar voice
(Marconi radio) Signal quantity, or
bandwidth in Hz
Distortion measured by Signal-
to-Noise ratio (SNR) in dB
devices, a second
voice DSP chip for
Noise source digitization
Sender Receiver and information
Digital Source
Modulation Channel Demodulation Decoding
compression is used.
Transmission Encoding
Lanwave provides
(Morse code radio) Signal in Bits/sec Signal distortion
measured in BER both chips and their
Transform using Inverse transformation from
collateral software to
a Coding Scheme knowledge of coding scheme our customers.

LanwaveTechnology, Inc. © 2001 As the consumer 2

information sources
increasingly turn digital, and device applications increasingly tied to the Internet, digital wireless
transmission is becoming necessary, natural and more popular every day, which is not for its performance
benefit over analog transmission alone.

The component count, manufacturing cost and engineering, are vastly different in between them. The
dichotomy is illustrated by the fact that technology suppliers in this market are largely different and besides
solutions from Europe, tends to offer unique and open architecture not yet standardized by committees.
However, the consumers and the product channels are the same. So how these digital wireless products shall
take advantage of new technology in new solutions to help the consumer; and how they are engineered,
manufactured and their benefits told through the channel; become pivotal to the delight and satisfaction of
all parties in the supply chain.

So what does digital transmission offer that analog don’t? And how do these new abilities bring better
solutions to the consumer? Summarized below are 7 unique digital advantages classified under the
dictionary definition of Personal, Communication and Networking:

Personal:
1. Encrypted for exclusive, secured communication.
2. Authenticated logon preventing interception faults.

Communication:
3. Natural and interactive, full duplex in between parties.
4. Multi lingual access with voice and data, combined to allow for future service upgrades.
5. Supporting circuit and packet connection and connectionless links to the backbone environment.

Network
6. STAR and BUS topology, flexibly supporting point-to-point and point-to-multipoint networking.
7. Open architecture protocol, or system command and control, easily configurable and customizable by the
consumer solution designer.

Lanwave © 2001 Page 2

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Spread Spectrum Coding

Digital CDMA improves performance over analog transmission

P1(f)

P2(f)

P3(f) N(f)

Bandwidth W

•Capacity-maximizing signal density functions Pi(f) over available

bandwidth W, as a function of frequency f, is the “filling” of the noise density function N(f), such as the situation
under uniform gravity. Total volume of the liquid is proportional to the available transmitter power.

•Maximum data rate (bit/s): C = W * log2(1 + S/N)

is proportional to signal bandwidth, but logarithmic to signal power. (Shannon et.al, c.1950.)

LanwaveTechnology, Inc. © 2001 3

It has been shown from research, that when the base band DSP chip spreads its signal over a wider
bandwidth using the correct ways and means, the resultant transmission performance would be vastly
improved over analog and narrow band methods. (C.E. Shannon et. al., since 1950.)

Performance in these cases is measured by the range, security, clarity and quantity of information
transmitted and received.

The practice of spread spectrum communication in military and aerospace applications has over 50 years of
history amongst first world countries. But until 10-years ago this technology was confined by the US
Government away from commercial use. Planetary satellites in the transmission of images employed the
technology over astronomical distances, while military radar and radios use anti-jamming and anti-
eavesdropping communications in the battlefield. Henceforth the ability and robustness of this technology is
well proven in non-cost sensitive applications. The challenges today are to commercialize the technology
for the consumer electronics market worldwide.

Lanwave © 2001 Page 3

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

CDMA history

Semaphore (500 years)

•smoke signal
•low bandwidth channel
•long range, low cost
•many drawbacks

Morse Code (100 years)

•EM wave alphabet of dots and dashes
•low implementation cost
•binary amplitude achieve a high noise margin
•high tolerance to phase, time & amplitude jitters
•short code reduces error correction ability

LanwaveTechnology, Inc. © 2001 4

The simplest ways to spreading a signal is by substituting the source signal with a more complex string
(phase chopping with a well chosen sequence of code). It is called direct sequence spreading or DSS. This
concept has been used for over 500 years in many cultures and situations from the early Indian semaphore
to the design of Morse code 100 years ago. Other methods include frequency hopping of the signal around
the applicable spectrum. This is the FH-TDMA technique used in European cordless telephone and similar
standards.

The earliest means of communication always employ a binary sequence, which tells of the reliability and
low requirement of the 1-bit architecture on the fundamental raw material, and therefore solution cost. This
binary architecture is still the most suitable for consumer electronics today. It can be shown mathematically
that 1-bit quantization may perform as well as other high-resolution methods, and in so doing leaves more
room to reducing external RF circuit costs.

Lanwave © 2001 Page 4

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Direct Sequence Spreading

DSS has a 50-year history

DSS works like a 2-tone semaphore

1 = Substitute with code “1101001” (e.g. RED)
0 = The inverse code “0010110” (WHITE)

..1101001 0010110 0010110 1101001 1101001..

1 0 0 1 1 0

LanwaveTechnology, Inc. © 2001 5

Borrowing the semaphore idea, direct sequence spreading or DSS, can be viewed as the substitution of
signal bits by a binary random sequence just long enough to satisfy FCC spectrum requirement to qualify
for the use in the unlicensed frequency bands, and yet kept short to limit implementation cost. The most
widely used in first generation DSS cordless telephones today employ a 12-bit sequence. The US FCC
commercial requirement can be implemented by 10X spread. But due to mathematical reasons the prime
number 11 is chosen, and the last 12th bit is used to balancing dc easing RF physical implementation.

Most DSS chips are of relatively low complexity by today’s digital VLSI standards, and therefore DSS
phones are able to approach analog telephone costs. However, DSS will not likely match analog raw
material cost since a super set of circuit and a higher complexity is being used, absent any business
motivated subsidy into the picture.

Lanwave © 2001 Page 5

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Code Division/Spread Spectrum

CD/SS™ is 1-bit CDMA for consumer electronics (5 years new)

CD/SS transmits a rainbow of color

•Increases signal density (bit/symbol)
•Allows longer chip sequence & powerful
error correction codes (Processing Gain)
•Signal density, coding gain are the sources
of higher performance over DSS.

..110100110111000111101011000111100010101..

110 011 000 111 101 100 000

CD/SS™ is a registered trademark by Lanwave Technology, Inc.
LanwaveTechnology, Inc. © 2001 6

In the CD/SS architecture Lanwave engineers employed a proprietary 1-bit DSP core to keep down cost,
and used CDMA techniques to improve performance:

1. Enhance range and clarity with a longer code than DSS (32-bit in the original SATURN chipset version,
to 192-bit equivalent in the latest SATURN-II DX family). A long code provides for better error correction
ability in a noisy environment. The ability to make correct (color) symbol decisions under random noise is
measured in a quantity called Processing Gain. The performance difference between CD/SS and DSS is in
between 4 to 10 decibel (dB), before “Link Layer” processing. This roughly translates into a doubling in
radio distance when the same class of radio component and transmission power is used. Longer codes also
render the communication to be more secure from eavesdropping.

2. Increase signal density by employing 8-level Code Division technique. In using an 8-level architecture
each CD/SS color symbol carries 3 signal bits increasing raw performance over DSS by 300%. However,
more coding levels will eventually introduce inter-symbol interferences, and will multiply DSP circuit
complexity exponentially. 8-level Code Division was chosen for its optimal balance in between silicon cost
and performance using today’s VLSI technology.

3. Adding embedded Link Layer processing on chip. This enabled a low software complexity. Externally
the Lanwave chip set typically requires a 4 to 8-bit class, 4K to 8K ROM, simple micro controller. It is
made possible by fully integrating all necessary physical and link layer routines in the on-chip micro code.
With embedded software the Lanwave CD/SS chip eases software development and shorten cycle time for
subsequent upgrades.

Lanwave © 2001 Page 6

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Framing allows for Multiple Access

Do you see the “S” in this picture?

there is an “S” in yellow, an “N” in light gray.

When viewing through a pure red filter, the blue
and green turns into black reviewing an arrow
pointing NE at the one o’clock direction.

LanwaveTechnology, Inc. © 2001 7

The CD/SS color symbols are framed to allow multiple signal messages to communicate within the same
spectrum, under external micro controller software control.

Lanwave © 2001 Page 7

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Wireless PBX, Wireless Internet,

Digital 2-way radio, Wireless Local Loop.. sharing cordless phone economy of scale

Retail / FOB ratio More features =

Code Division
higher added value Spread Spectrum
(CD/SS)

Direct Sequence
Spreading (DSS)

1 base 1 handset
with extended range
CT0, CT1, Analog
cordless phone
1 base 1 handset

Performance & Features

LanwaveTechnology, Inc. © 2001 8

Multiple Access ability enables a cordless telephone designer to readily develop more sophisticated, higher
value added products such as wireless PBX, wireless Internet and wireless local loop systems that are more
valuable at retail and enjoying a higher manufacturing and distribution margin. Fast upgrade ability enables
manufacturers and brand providers to meet the demand of consumers from the home, SOHO to industrial
and rural telecommunication market segments. Easy customization allows system manufacturers to build
unique, architecture secure and therefore defensible business models and products.

Lanwave © 2001 Page 8

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Signal density & RF radio range

Advantage over Direct Sequence Spreading (DSS)

Direct Sequence Spreading Code Division/Spread Spectrum

(DSS) (CD/SS & CD/SS with VPSK**)

time
time

0 blue 001
F(/pn) 0 red 010
F(pn) 1 green 111

frequency frequency

* F(pn) is the Discrete Fourier Transform (DFT) of the random code pn.
** VPSK (Very Minimal Phase Shift Keying) is adding Pulse Width Modulation on Morse Codes to increase signal density.

LanwaveTechnology, Inc. © 2001 9

Technically comparing CD/SS with DSS technique:

Empirically CD/SS achieved a 300% signal density over DSS. When a new technique incorporating pulse
width modulation is employed (supported in second and third generation Lanwave chip sets), it further
improves bandwidth efficiency by an additional factor of 2 (SATURN VX2) to 6 (SATURN DX2). Next
generation SATURN chipsets may further increase this factor from the improvement in semiconductor
processing technology.

Higher signal density, more error correction and efficient spectrum utilization combined to translate into
better radio (RF) sensitivity. The benefit is a longer wireless range for the same transmission power; or
reduced power consumption for the same wireless range. Under Link Layer software control these benefits
extends usable range and talk time while keep down audio humming. All are direct comparison items in the
retail telephone channel, and driving factors for future consumer wireless devices.

Lanwave © 2001 Page 9

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Fast Recovery & Synchronization

Advantage over DECT and Frequency Hopping with TDMA

Frequency Hopping Spread Spectrum Code Division/Spread Spectrum

(2.4GHz DECT with FH-TDMA) (CD/SS with VPSK)

time Recover time

Black Out
Sequence
Sequence
Broken
Broken

frequency frequency

• Long “BLACK OUT” on FCC required hopping sequences

• Reduces performance exponentially as environment become noisy

LanwaveTechnology, Inc. © 2001 10

DECT (short for Digital European Cordless Telephone) employs narrow band, frequency hopping (FH-
TDMA) to meet US FCC spectrum requirement in the 2.4GHz range.

Frequency hopping technique is quite different from DSS or CD/SS in many ways. One in particular is that
it requires a long recovery period when two communicating parties have lost touch. The process is similar
to finding a lost person inside a crowded shopping mall, when both agreed that one would search counter
clockwise, and the other clockwise, through each shop. Eventually both parties will meet but this
synchronization process may take a long time.

CD/SS technique does not suffer from this problem since its receiver scans every frequency simultaneously
(like having a detector inside every shop in this example) and can immediately locate its partner in the next
search period. This short recovery ensures that the “Black Out” during which no reception is recorded from
the DSP chip can be kept down to a very short duration.

The synchronization problem of FH-TDMA is not limited to DECT, but in all frequency hopping solutions.
This problem escalates as the environment becomes more crowded and noisy in the near future.

One other advantage of CD/SS over FH-TDMA is in better RF power tuning. This feature enables a CD/SS
DSP chip to calculate the physical and radio distance with its partner, thereby reduces RF transmission
power when allowed, and increases only as necessary. It has far-reaching performance implications in
improving spectrum capacity and reducing inter system jamming. CD/SS architecture provides a natural
facility for the micro controllers to implement a closed loop tuning solution.

Lanwave © 2001 Page 10

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Simplified Code Overlapping

Advantage over traditional CDMA and W-CDMA

Code Division with Multiple Access Synchronous CD/SS

(CDMA & W-CDMA) (S-CD/SS with VPSK)

time time

Time of Arrival
(TOA) varies
from the lack
of source timing
synchronization

frequency frequency

• Wide variation in “TOA” increases DSP complexity and cost

• S-CD/SS controls code overlap without sacrificing performance

LanwaveTechnology, Inc. © 2001 11

Technically comparing CD/SS with wide band CDMA, or W-CDMA technique:

Wide band CDMA (W-CDMA) is employed in next generation mobile phone systems. Its advantage is
particular to the high-speed roaming requirement of mobile phones.

Because signal arrival time and energy at the receiving antenna is a function of frequency and its interaction
with environment objects, a wide band signal has a wider time variation due to multi path effects and
Raleigh fading. A special circuit called Rake equalizer is used to compensate for the differential arrival time
in the base band DSP. In addition, a packet speech processor is generally employed to allow bursty
interruptions from the Rake receiver. Both circuits are not needed in CD/SS architecture. The benefit of
code overlap is taken advantage in a synchronous CD/SS design that coordinates transmit timing on
multiple code signals.

The reduction in circuit complexity, combined with features such as 1-bit RF quantization reduces the cost
of a CD/SS handset to less than one-third of a comparable mobile phone handset. (Less than $20 compared
to approximately $60.) And in a typical rural wireless local loop environment the advantage in Rake
receiver compensation is not needed, and hardly justifiable by a 3 times increase in unit cost.

Lanwave © 2001 Page 11

 20111 Stevens Creek Blvd., #260
Cupertino, CA 95014. U.S.A.
Tel: 408.253.3883
Fax: 408.253.6630

Spectrum Performance
Hertz/bit/second Coding Gain
============= ==========
DSS (12-chip, BPSK) 18 Hz / b/s 11 dB

CD/SS (GMSK, TDD full duplex)

- Gen 1 (VX), obsoleted. 32 Hz / b/s 15 dB
- Gen 2 (VX2) 16 Hz / b/s 15 dB
- Gen 3 (DX2) 5 Hz / b/s 10 - 12 dB
G3 un-spread mode 3 Hz / b/s 0 dB
G3 un-spread burst mode 1.5 Hz/b/s 0 dB

Manchester Coding 6 Hz / b/s 0 dB

DECT /FH-TDMA 3 Hz / b/s 0 dB

Narrow Band Digital 3 - 6 Hz / b/s 0 dB

LanwaveTechnology, Inc. © 2001 12

These figures are calculated using the most typical RF implementation.

3rd Generation CD/SS chipset

Price, performance
Applications
L9002DX-40P L9002DX2-40 Packet voice, WLL,
& Wireless Internet

L9002DX-40P
L9002DX-20P
L9002DX2-33 PCM voice transfer
L9002VX DX2
L9002DX2-20 ADPCM voice over data
VX2 *
L9002VX2 / L9320 MBMH & 2-way Radio

Oct., 97 Dec., 98 Jan., 2000

LanwaveTechnology, Inc. © 2001 13

All SATURN chip families are backward software compatible at the binary code level.
This protects customer software investment and allows short R&D cycle time after the
initial CD/SS product has been engineered and introduced.

-- END --

Lanwave © 2001 Page 12