Electrical Engineering Dictionary - part 159

put load of the following stage, in order to
compensate for high frequency loss due to
the distributed capacitance of the two stages.

shunt reactor

a reactor intended for con-

nection in shunt to an electric system to draw
inductive current.

periodic table symbol for silicon. See

silicon

Si/SiGe/SiGeC

silicon-based alloy sys-

tem providing band offsets that enable het-
erostructures that can be utilized for hetero-
junction transistor design and quantum con-
finement.

side lobe level

the ratio of a local maxi-

mum in a radiation pattern to the global max-
imum (main beam) of the radiation pattern.

sideband

the signal produced when a car-

rier signal is modulated. They may be one
single sideband, one set of upper and lower
sidebands, or a series of sidebands whose
number is dependent on the modulation in-
dex of the modulation system being used.

sidelobe

a lobe in an antenna radiation pat-

tern apart from the main lobe and any grating
lobes. Sidelobes have peak amplitudes less
than that of the main lobe.

sidelobe level (SLL)

the peak amplitude

of a sidelobe relative to the peak amplitude of
the main lobe. The SLL is usually expressed
as the number of decibels below the main
lobe peak.

Siemens, Ernst Werner von

(1816–

1892) Born: Lenthe, Hanover, Germany

Siemens is best known for the German

and British companies that bear his name.
Siemens was a strong believer in basic re-
search, as well as an avid inventor.

His

early inventions included an improved gutta-
percha wrapped telegraph cable that allowed
his companies to secure a number of lucrative
cable contracts. His discovery of the dynamo

principle, and his use of this in heavy-current
applications, allowed his companies to be-
come pioneers in devices to generate electric-
ity and rail applications. Siemens’ belief in
basic research made him a champion of stan-
dards and research institutions that he helped
to establish.

SiGe

in order to increase the speed of Si

semiconductor devices without compromis-
ing on Si’s ease of device processing, SiGe
heterostructures can be used.

sigma-delta A/D conversion

an over-

sampling A/D conversion process where
the analog signal is sampled at rates much
higher (typically 64 times) than the sampling
rates that would be required with a Nyquist
coverter. Sigma-delta modulators integrate
the analog signal before performing the delta
modulation. The integral of the analog sig-
nal is encoded rather than the change in the
analog signal, as is the case for traditional
delta modulation. A digital sample rate re-
duction filter (also called a digital decimation
filter) is used to provide an output sampling
rate at twice the Nyquist frequency of the
signal. The overall result of oversampling
and digital sample rate reduction is greater
resolution and less distortion compared to a
Nyquist coverter process.

sigma-delta modulation

a method for

scalar quantization, similar in principle to
delta modulation but somewhat more sophis-
ticated.

Employed in, e.g., compact-disk

players.

sigmoid function

a compressive function

that maps inputs less than -1 to approximately
zero, inputs greater than 1 to approximately
1 and maps values from -1 to 1 into the range
0 to 1. A common sigmoid function is

+ e

−

Sigmoid functions are often used as acti-
vation functions in neural nets.

See also

activation function

2000 by CRC Press LLC

sigmoidal characteristic

a widely used

type of activation function, especially in net-
works trained using schemes like backprop-
agation that are based upon gradient descent.
The most common functions used are the arc-
tan, tanh, and logistic functions, with appro-
priate variations for binary and bipolar vari-
ables.

sign flag

a bit in the condition code register

that indicates whether the numeric result of
the execution of an instruction is positive or
negative (1 for negative, 0 for positive).

sign-magnitude representation

a num-

ber representation that uses the most signifi-
cant bit of a register for the sign and the re-
maining bits for the magnitude of a binary
number.

signal

a real or complex function,

f (t) of a

time variable

t. If the domain is the real line

t ∈ R, or an interval of the real line, then
the signal is a continuous time signal. For
example,

f (t) could represent the magnitude

of a voltage applied to a circuit, as shown in
the figure.

Continuous time signal.

If the domain is a discrete set

{k, k =

, 1, 2, . . .}, then the signal is a discrete time

signal. Discrete time signals commonly arise
in engineering applications by sampling a
continuous time signal, in control and filter-
ing applications. For example,

f [k] could

represent the above voltage sampled every
0.5 seconds, as shown in the figure.

Discrete time signal.

signal averaging

an averaging process

that is used to enhance signals and suppress
noise, thereby improving the signal-to-noise
ratio. See also

averaging

signal conditioning

a process that is used

to improve the appearance or effectiveness
of a signal, either by suppressing noise or by
transforming the signal into a more suitable
form. This latter category includes image
enhancement. Signal conditioning is often
appropriate in preparing signals for recogni-
tion.

signal constellation

in digital communi-

cations, the set of transmitted symbols repre-
sented as points in Euclidean space. For ex-
ample, the signal constellation for (uncoded)
quadrature amplitude modulation is a set of
points in the complex plane.

signal decimation

See

decimation

signal detection

detecting the presence of

a signal in noise.

signal flow graph

graphical representa-

tion of the relationships between a set of in-
dependent input variables that are linearly re-
lated to a set of dependent output variables.

2000 by CRC Press LLC

signal level

the value of a signal at a par-

ticular place and time.

signal processing

a generic term that

refers to any technique that manipulates the
signal, including but not limited to signal
averaging, signal conditioning and signal
recognition. When applied to images, it is
normally referred to as image processing, the
term signal processing usually being reserved
for 1-D signals.

signal recognition

the recognition of sig-

nals by appropriate analysis, often with the
help of filters such as matched filters or fre-
quency domain filters.

signal recovery

the process of extraction

of signals from a background of noise or clut-
ter, often in situations, where the signal-to-
noise ratio is so low that matched filters, syn-
chronous detectors or lock-in amplifiers have
to be used. See also

synchronous detection

signal reference subsystem

this subsys-

tem provides the reference points for all sig-
nal grounding to control static charges, noise,
and interference. It may consist of any one
or a combination of the lower frequency net-
work, higher frequency network, or hybrid
signal reference network.

signal restoration

the restoring of data

that has been corrupted by instrumentation
dynamics and noise.

signal subspace

in an orthogonal de-

composition of a space, the part to which
the desired signal belongs. See also

noise

subspace

signal transfer point (STP)

a packet

switch found in the common-channel signal-
ing network; it is used to route signaling mes-
sages between network access nodes such as
switches and SCPs.

signal variance

See

variance

signal-to-interference ratio (SIR)

the ra-

tio of the average power of the signal compo-
nent to the average power of the interference
component in a case where an information-
bearing signal of interest has been corrupted
by interfering signals.

signal-to-noise plus interference ratio
(SNIR)

the ratio of total signal power to

the sum of total noise power and total inter-
ference power at a receiver. The SNIR is
a more complete indicator of received sig-
nal quality than either SIR or SNR, where
the relative contribution of receiver noise and
external sources of interference are either
unknown or widely varying.

It is a unit-

less quantity. See also

signal-to-noise ratio

signal-to-interference ratio

signal-to-noise ratio (SNR)

the ratio of

the average power of the information signal
component to the average power of the noise
component in a signal consisting of the sum
of an information signal component and a
corrupting noise component. It is a unitless
quantity.

signaling

procedures used to control (set

up and clear down) calls and connections
within a telecommunication network.

signaling system 7 (SS7)

a communica-

tions protocol used in common-channel sig-
naling networks.

signature

a characteristic easily com-

puted feature or function by which a particu-
lar object or signal may be at least tentatively
identified. An example is the centroidal pro-
file for an object having a well defined bound-
ary.

signature analysis

(1) a test where the

responses of a device over time are compared
to a characteristic value called a signature,
which is then compared to a known good one.

(2) an analysis of the signature to extract

the desired (signal) information.

2000 by CRC Press LLC

signed-digit representation

a fixed-radix

number system in which each digit has a sign
(positive or negative). In a binary signed-
digit representation, each digit can assume
one of the values

−1, 0 and 1.

significand

the mantissa portion of a

floating-point number in the IEEE 754
floating-point standard. It consists of an im-
plicit or explicit leading integer bit and a frac-
tion.

signum function

the function

sgn

(t) =






+1 t > 0
0

t = 0

−1 t < 0.

Used in modeling numerous types of system
functions.

SIL

See

surge impedance loading

silicon

most common element in the

earth’s crust and a type IV (from periodic
table of elements) semiconductor material.
The bipolar carriers, both holes (p dopants) or
electrons (n dopants) are roughly in propor-
tion to each other, resulting in nearly equal
currents in the same direction. They move
at about half the speed of comparable GaAs
unipolar carriers. The thermal resistance is
also about half that of GaAs.

Silicon has an indirect band gap of 1.11

eV, density-of-states masses of 1.1 times the
free-electron mass for the conduction band
and 0.56 for the valence band.

silicon compiler

a set of software pro-

grams intended to start with design equations
and output the corresponding GDS2 data.
Silicon compilers are currently used to trans-
late a standard cell design from one set of
design rules to another or to create a new set
of standard cells.

silicon controlled rectifier (SCR)

a cur-

rent controlled four-layer (pnpn) device for
high power (3000 A) and low speed (500
Hz) applications. SCRs can only be on or

off, with no intermediate operating states like
transistors. The SCR acts as a switch that
is turned on by a short current pulse to the
gate, provided that the device is in its forward
blocking state. Once latched on, the gate cur-
rent can be removed and the device will re-
main on until the anode current goes negative,
or the current through the SCR becomes less
than its designated holding current. A disad-
vantage is that a commutation circuit is often
needed for forced turn-off (forced commuta-
tion).

silicon dioxide (SiO

)

an amorphous

compound of silicon and oxygen with a resis-
tivity of 10

to 10

ohm-cm and a bandgap

of 8.1 eV. Essential as a dielectric and insu-
lator for silicon devices.

silicon nitride (SiN

)

for a semiconductor

manufacturing process, a compound formed
of silicon and nitrogen that is deposited and
etched back to provide a masking layer ap-
propriate for withstanding subsequent high
temperature processing such as oxidation.

SIMD

See

single-instruction stream,

multiple-data stream

similarity between symbol strings

See

distance between symbol strings

similarity measure

the reciprocal con-

cept of distance measure.

See

distance