Innovation
through
Mathematics
 
     
 

Home
About Us
Products
Technology
  What is SDM?
  Cable Apps
  SDM University

   White Papers &
    Presentations
  Patents
  Archives
News / Events
Careers
Contact Us
Feedback
Site Map

 

 

 

Privacy Statement
Copyright Statement

 
 


What is SDM?

 

Overview

Sub-band Division Multiplexing (SDM) is a technique of dividing RF spectrum into multiple, equal-sized sub-bands:


Click on Image for Larger Picture

Unique to SDM, each sub-band in a multi-sub-band system has the properties of:

  • 50dB or more stopband (tunable)
  • Orthogonal in both time and frequency
  • Very high frequency division independence

The mathematics used at Broadband Physics are based on the novel application of Wavelet Mathematics for information signaling. The underlying principle of using Wavelet math produces an inherent multi-sub-band system. The transmitter for each sub-band is a filter which takes digital data and spreads the information over a Wavelet Impulse. Each sub-band receiver is an equivalent filter which converts the Wavelet Impulse back to digital data.

With SDM Wavelet mathematics, a filter is used per sub-band to generate the transmission Wavelet Impulse signal and a corresponding filter is used to decode the impulse.

Each sub-band operates in the same manner, but with an offset frequency which is orthogonally related to the other sub-bands. One signaling symbol is comprised of the sum of the Wavelet Impulses from all active sub-bands. At the receiver, each sub-band is initially recovered by use of a Finite Impulse Response (FIR) filter. Each sub-band's FIR filter can be constructed from digital logic or through the use of a passive Surface Acoustic Wave (SAW) filter. From there, a digital equalizer and other logic reconstructs the original data. SDM is a multi-band carrier-less transmission system.

The SDM Class of Modulation

SDM represents a leap in modulation technology by creating a new class of transmission based on multi-sub-band carrier-less impulse/filter signal transmission.


Click on Image for Larger Picture

There are several distinct variations of SDM:

  • SDM. Baseband processed multi-band impulse/filter. Suited for both baseband and RF applications.
  • coSDM (coded). Combines SDM with spread coding techniques to improve robustness in certain application environments.
  • fmSDM (frequency modulated). A variant of SDM which creates a constant envelope RF carrier containing a special set of Wavelet Impulses. Trading data capacity for power, fmSDM permits the use of class C amplifiers resulting in an 85% to 90% savings in power consumption compared to existing wireless LAN methods.
  • uwSDM (ultra wide). A variant of SDM where an ultra wide RF passband is divided into very wide sub-bands. uwSDM inherits all the properties of SDM, e.g. high efficiency, robustness, software controlled notching, etc.

Each of the SDM variants can be used in any of the existing last mile and last yard access network and WAN/LAN techniques.

Broadband Physics' first patent was issued in 1994. It and our subsequent patents cover the multi-sub-band class of modulation. Broadband Physics is committed to standardize SDM for applications that select SDM as the preferred signal transmission method.

 

SDM Compared to Single-Carrier and Single-Band

In the class of "single" modulations for last mile access networks, modulations break down into two distinct categories: single-carrier, and carrier-less single-band.


Click on Image for Larger Picture

Following the single-band single-carrier branches, there is a variety of PSK (Phase Shift Keying) techniques, VSB (Vestigial Side Band), QAM (Quadrature Amplitude Modulation) as well as the spreading techniques CMDA (Code Division Multiple Access) and S-CDMA (Synchronous Code Division Multiple Access). There are two carrier-less single-band modulations CAP (Carrier-less Amplitude/Phase) which is essentially a carrier-less equivalent to QAM, and UWB (Ultra Wide [single] Band) which has gained notoriety for promising to open up otherwise dedicated RF spectrum for unlicensed use.

SDM Compared to Carrier-less Single-Carrier

In the frequency domain, single-carrier modulations require the use of a Raised Cosine filter to shape the stopbands (guardbands) of the carrier. This is necessary to allow adjacent carrier operation on systems with multiple channels, such that one carrier does not interfere with its immediate neighbors. The use of Raised Cosine guard bands removes between 13% to 25% of the available RF spectrum from use. SDM, however, requires only one sub-band of guardband to be turned off between SDM and an adjacent channel. Dependent on the number of sub-bands provisioned, the frequency overhead could be as low as 1% overhead.

In the frequency domain, SDM uses a divide and conquer approach to sub-divide the RF passband into smaller, more manageable allocations. This allows SDM to be much more robust in the presence of certain types of noise impairments such as phase ripple, channel tilt/twist, micro-reflections, and multi-path distortions.

In the time domain, the nature of Wavelet Impulses allows a much less complicated equalizer at the receiver. This reduces complexity and implementation costs. SDM symbols are longer in time than the corresponding symbols in single-carrier modulation. This allows SDM to be tuned to be relatively insensitive to known time impairments such as micro-reflections in the home on cable. Single-carrier modulations can require a very sophisticated and complicated adaptive Time Domain Equalizer (TDE) to cope with channel impairments, whereas SDM produces superior results with much less complexity.

SDM Compared to Carrier-less Single-band

Broadband Physics technology is much more efficient that other existing carrier-less single-band modulations (e.g. CAP Carrier-less Amplitude/Phase, UWB Ultra-Wide Band, multi-band UWB) as SDM simultaneously transmits an impulse on each sub-band and, in addition, SDM transmission symbols overlap in time due to the time orthogonality properties of Wavelet mathematics, thereby increasing the digital data capacity of the channel.

 

SDM Compared to Multi-Carrier

Another class of modulation is called multi-carrier, which includes techniques such as DMT (Discrete Multi-Tone) and OFDM (Orthogonal Frequency Division Multiplexing).


Click on Image for Larger Picture

Multi-carrier modulations uses multiple equal-sized carriers (tones)to fill available spectrum. Each tone is essentially a single-carrier QAM signal, but without the Raised Cosine bandpass filter which is required to contain the stopband energy to a single RF channel. With DMT/OFDM, the stopband (sideband) energy of interior tones is spread across many adjacent tones; data is recovered through the properties of frequency orthogonality. This has the unfortunate consequence that narrow-band interference effects many overlapping tones and that phase ripple requires both frequency and time domain equalization, which is an increasingly difficult computational algorithm as multi-carrier modulations are pushed into more and more application spaces. In addition, OFDM requires the use of a time-domain overhead called a Cyclic Prefix. The Cyclic Prefix is required to overcome issues of phase ripple, micro-reflections, and multi-path distortions. Typically, the time based overhead of the Cyclic Prefix is on the order of 2% to 5% of the digital capacity of the system, but can be as high as 30% to 40% in certain application environments. SDM requires neither a Raised Cosine filter, nor a time domain equalizer, nor a Cycle Prefix.

In comparison to single-carrier and multi-carrier systems, SDM does not require a Raised Cosine filter, does not require the use of a complex Time Domain Equalizer (TDE), nor the use of a Cyclic Prefix.

The steep 50dB stopband directly results from the use of Wavelet Mathematics and the creation of signaling symbols by use of a filter. The parameters of the filter can be adjusted to produce a stopband of any size allowing the tailoring of the system for different applications. The steep inherent stopband directly results in the signaling power falling off completely within about 1/2 a sub-band width into the immediate neighbors. This has several immediate benefits:

  • Very high adjacent sub-band independence: the receiver equalization complexity is dramatically reduced as at worst case, the equalization spectrum is the sub-band in question and the entire width of its two immediate neighbors.
  • The width of the sub-band can be tuned to virtually ignore Phase Ripple impairments on a application by application basis. Phase Ripple is more commonly known as micro-reflections in cable, and multi-path in wireless.

 

Comparison Table

The following table summarizes the key distinctions between SDM and multi-carrier modulation signaling transmission systems:

Notes:

  1. Carriers/Tones overlap over many side tones due to fixed 13dB stopband of Fourier Transform mathematics compared to tune-able 50+dB stopband of Wavelet filter mathematics.
  2. A TDE is required to attempt to reconstruct effects of real-world non-linear channel impairments for QAM and DMT/OFDM demodulation mathematics. TDE’s are complex mathematical “engines” that become increasingly more complex as QAM/DMT/OFDM are pushed for more performance. It is difficult for a TDE to reconstruct all non-linearity effects.
  3. SDM math inherently copes with real-world non-linearities, therefore no TDE required
  4. Time domain overhead (3% - 40%) of Cyclic Prefix required to wait for impulse transients to “die down” between modulation symbols
  5. SDM is and impulse/filter signaling technique; therefore no Cyclic Prefix required
  6. Raised Cosine filter not required for SDM

Software Defined Notches

In SDM, not all sub-bands need be "turned on". Turning off two adjacent sub-bands creates a 50dB notch at any desired band within the spectrum of operation; this is done without the use of external hardware notch filters. Software control is used to provision the system, allowing notches to be created or moved on demand.


Click on Image for Larger Picture

With SDM, the sub-band widths are sized for the application space. Software controlled notches are very useful for all last mile access networks as well as Local Area Network (LAN) systems, such as:

  • Cable
  • Twisted-pair networking, last mile and in the home/office
  • Power Line Communications: last mile and in the home/office
  • Ultra Wide Band (UWB)
  • Wireless: WAN, LAN and point-to-point
  • Broadcast

With software control of sub-bands, both passbands and notches can be provisioned at deployment time to meet any regulatory requirements. Changes in regulations can be accommodated with a simple a software change rather than a costly hardware change.

Summary

Generally speaking SDM better handles real world effects of non-linear impairments found in most Last Mile access networks as compared to QAM, DMT, and OFDM. SDM inherently does not require frequency or time based overheads as compared to QAM, DMT, and OFDM. SDM's major strengths apply to signal transmission environments where digital capacity is challenged by real world non-linear impairments and where software defined notches and passbands create new opportunities for increasing broadband capacity.

SDM inherently is a less complex method of obtaining high-capacity, high-efficiency, and high-robustness and maintains a high degree of flexibility for provisioning dynamically and as well as at initial time of deployment.