# Statistical Distributions in Telecommunications

Mobile Comms. Systems

1

Basic Notions
BaNo (1/4)

• The use of statistical models is essential to describe: • non-guided propagation in random environments; • user’s mobility; • phone calls and data connections; • users’ influence in network performance.

Mobile Comms. Systems

2

Basic Notions
BaNo (2/4)

• Given the probability density, p(x), and cumulative distribution functions, P(x), i.e., PDF and CDF, one has dP p ( x) = dx or x P( x) = ∫ p (t ) dt where
−∞

P(x) = Prob(X ≤ x) • One has
Mobile Comms. Systems

Prob(|X| ≤ x) = P(x) - P(-x)

3

x 2 ∞ −∞ 2 x = ∫ x 2 p ( x) dx • median. xm xm Mobile Comms.Basic Notions BaNo (3/4) • The main parameters are: • average. x x = ∫ x p ( x) dx −∞ ∞ • mean square. Systems −∞ ∫ p ( x) dx = ∫ p ( x) dx = 1 / 2 xm ∞ 4 .

σx2. μn ∞ −∞ μ n = ∫ (x − x ) p ( x) dx n Mobile Comms. mx p(mx) = max[p(x)] • moments.Basic Notions BaNo (4/4) • mode. σx. and standard deviation. Systems • variance. σx2 = μ2 • Chebyshev’s inequality allows to quantify the dispersion of the random variable: 1 Prob( x − x ≥ kσ x ) ≤ 2 k 5 .

Systems 6 . ⎩ ⎧ 0 . x<a ⎪x − a ⎪ P( x) = ⎨ . a≤ x≤b ⎪ p ( x ) = ⎨b − a ⎪ 0 . b<x ⎩ Mobile Comms. o.a ≤ x≤b ⎪b − a ⎪ 1 .c.Uniform Distribution UnDi (1/2) • CDF and PDF: ⎧ 1 .

Mobile Comms.Uniform Distribution UnDi (2/2) • Parameters: a+b x = xm = 2 b−a σx = 12 a 2 + ab + b 2 2 x = 3 • It is used to describe the phase of a signal. Systems 7 .

Normal (Gauss) Distribution NoDi (1/5) • PDF and CDF: − 1 p ( x) = e 2π σ x ( ) . x ∈ IR x− x 2σ x 2 ⎛ x−x ⎞ 1 + erf ⎜ ⎟ ⎝ 2σ x ⎠ P( x) = 2 • Parameters: x = xm = mx Mobile Comms. Systems 2 x = x +σx 2 2 8 .

Mobile Comms. Alternatively. v>0 9 .661u + 0.339 u + 5. Systems e erf (v) ≅ 1 − π (π − 1)v + v 2 + π −v 2 .51 ( −u /2 2 ) where x−x u= σx for u > 0.Normal (Gauss) Distribution NoDi (2/5) • Approximation for P(x) e P(u ) ≅ 1 − 2 2π 0.

282 2.167⋅10-5 2. Systems 1 .275⋅10-2 1.P(u) 5.867⋅10-7 u 1.587⋅10-1 2.326 3.719 4.P(u) 10-1 10-2 10-3 10-4 10-5 10-6 5 10 .753 1 .090 3.265 4.Normal (Gauss) Distribution NoDi (3/5) • Occurrence intervals u 0 1 2 3 4 Mobile Comms.350⋅10-3 3.000⋅10-1 1.

Systems 11 [Source: Boithias. 1983] .Normal (Gauss) Distribution NoDi (4/5) • Diagrams in Gauss Scale Mobile Comms.

Mobile Comms.Normal (Gauss) Distribution NoDi (5/5) • It is used to describe fluctuations around a mean value. Systems 12 . • It cannot be used to describe entities that cannot be negative.

Systems 13 . x>0 ⎛ ln( x) − xl ⎞ 1 + erf ⎜ ⎟ 2σ l ⎠ ⎝ P( x) = 2 Mobile Comms.Log-Normal Distribution LNDi (1/3) • PDF e CDF: 1 1 p ( x) = e 2π σ l x ⎛ ln( x ) − xl ⎞ −⎜ ⎟ ⎜ 2σ ⎟ ⎝ l ⎠ 2 .

Systems σl 2 −1 e xl +σ 2 / 2 l • It is used to describe entities signal power or amplitude. 14 .Log-Normal Distribution LNDi (2/3) • Parameters: x=e 2 xl +σ 2 / 2 l x =e 2 xl +σ 2 l ( ) xm = e xl mx = e xl −σ 2 l σx = e Mobile Comms. namely slow fading.

Vol. V.Log-Normal Distribution LNDi (3/3) • PDF and CDF: normal log-normal Mobile Comms. 1007] . Rep. Systems 15 [Source: ITU-R.

Systems σ x = (4 − π ) x 2 2 16 .Rayleigh Distribution RaDi (1/3) • PDF and CDF: 2 x -x2 / x2 p ( x) = 2 e . x>0 x P( x) = 1 − e -x2 / x2 • Parameters: x = π x2 2 mx = x 2 2 xm = ln(2) x 2 Mobile Comms.

Rep. Vol. Systems 17 [Source: ITU-R.Rayleigh Distribution RaDi (2/3) • PDF and CDF: Mobile Comms. V. 1007] .

• It is used to describe intense fast fading.Rayleigh Distribution RaDi (3/3) • It is associated to the magnitude of the sum of vectors having amplitudes with a normal distribution and phases with a uniform one. Systems 18 . Mobile Comms.

Suzuki Distribution SuDi (1/2) • PDF and CDF: x − x2 / v2 p R ( x. Systems 19 . Mobile Comms. v ) = 2 e v /2 pLN (ν ) = 1 1 2π σ l ν e ⎛ ln( ν ) −νl ⎞ −⎜ ⎟ ⎜ 2σ ⎟ ⎝ l ⎠ 2 1 P( x) = 1 − ∫e 2π −∞ − x 2 e − 2 σ l t −t 2 / 2 +∞ dt (vl = 0) • It describes joint slow and fast fading.

Rep.Suzuki Distribution SuDi (2/2) • CDF: x / x2 σl Mobile Comms. V. Systems 20 [Source: ITU-R. Vol. 1007] .

Systems 21 .Rice Distribution RiDi (1/3) • PDF: 2 2 x −(x 2 + x02 ) xR ⎛ x x0 ⎞ p ( x) = 2 e I0 ⎜ 2 ⎟ ⎜ x /2⎟ xR ⎝ R ⎠ describing the sum of a fixed vector (amplitude x0) 2 with a Rayleigh distributed vector (power xR ). • It is used to describe non-intense fast fading. • Usually. the Rice parameter is used 2 2 K[ dB] = 10 log x0 xR ( ) Mobile Comms.

Systems 2 xR 2 x0 22 [Source: Parsons. 1992] .Rice Distribution RiDi (2/3) • PDF: K→0 K≈1 K>>1 Mobile Comms.

1007] . Systems 23 [Source: ITU-R.Rice Distribution RiDi (3/3) • CDF: x 2 2 x0 + xR 2 xR κR = 2 2 x0 + xR Mobile Comms. Vol. V. Rep.

Exponential Distribution ExDi (1/2) • PDF and CDF: 1 -x / x p ( x) = e . Systems 2 σ =x 2 x 2 24 . x>0 x P( x) = 1 − e -x / x • Parameters: xm = x ln(2) mx = 0 x =2x 2 Mobile Comms.

Mobile Comms.Exponential Distribution ExDi (2/2) • It is used to describe the duration of various phenomena. namely associated to signal fading and phone calls. Systems 25 .

Mobile Comms. • Parameters are s=q 2 σ s = q (1 − q ) • It is used to describe the occupation of a telecommunications channel. Systems 26 . 1 where q is the probability of occurring 1.Bernouli Distribution BeDi (1/1) • Mass probability function s s p ( s ) = q (1 − q ) . s = 0.

k = 0. Systems 2 σ k = n q (1 − q ) 27 .Binomial Distribution BiDi (1/2) • Mass probability function: ⎛n⎞ k n−k p (k ) = ⎜ ⎟ q (1 − q ) . n ⎝k ⎠ where q is the occurrence probability for each of the n times. • Parameters k =q mk = int[(n + 1)] q Mobile Comms. K.

Mobile Comms.Binomial Distribution BiDi (2/2) • It is used to describe phone calls. where q = λt/n. Systems 28 . t being the sampling time interval and λ the average phone calls generation.

α < 1 mk = ⎨ ⎩α . Systems 29 . Mobile Comms. α >1 • It is used to describe the generation of phone calls (α = λt).Poisson Distribution PoDi (1/1) • Mass probability function: k α −α p(k ) = e k! • Parameters: k =α σ k2 = α ⎧0 .

Exponential Distribution BeDi .Bernouli Distribution BiDi .Log-Normal Distribution RaDi .Suzuki Distribution RiDi .Basic Notions UnDi .Uniform Distribution NoDi .Rice Distribution ExDi .Rayleigh Distribution SuDi .Normal Distribution LNDi . Systems 30 • • • • • • • • • • • BaNo .Binomial Distribution PoDi – Poisson distribution .Table of Contents ToC (1/1) Mobile Comms.