Skip to main content

S and T switches; 3-stage vs 2-stage

Short answer

Space switches change highway, time switches change time slot. Two-stage TS/ST networks offer a single path, so blocking is high. TST uses an internal slot x: input TSI moves to x, space connects highways in x, output TSI to destination slot, giving l paths. STS uses k centre TSIs, providing k paths. Blocking = [1-(1-p')^2]^k or ^l, vastly reduced.

What do you mean by S (space) and T (time) switches? Show that 3-stage STS or TST network can minimize the switching problems associated with 2-stage ST or TS network with their working models.

S and T switches

  • Space (S) switch: a time-multiplexed crosspoint matrix that connects an input TDM highway to an output highway in a given time slot. It changes the highway but not the slot.
  • Time (T) switch / TSI: a speech memory with a control memory that writes samples and reads them in a different order. It changes the slot but not the highway.

Problem with two-stage ST and TS networks

To connect slot ii on highway aa to slot jj on highway bb, both the slot and the highway must change.

 TS: HWa --[T: i->j]--[S: a->b in slot j]-- HWb
 ST: HWa --[S: a->b in slot i]--[T: i->j]-- HWb
  • In TS, the space stage must use slot jj; if slot jj is already used by another input highway to reach bb, the call is blocked.
  • In ST, the space stage must use slot ii; if slot ii is busy into highway bb, the call is blocked.
  • There is only one path, so the blocking probability equals the chance that this particular link is busy (about pp). This is too high for an exchange.

Three-stage TST network

 HWa->[TSI-A: i->x]->[S: a->b in x]->[TSI-B: x->j]->HWb
  1. Input TSI moves the sample from slot ii to any free internal slot xx.
  2. Space stage connects aa to bb in slot xx.
  3. Output TSI moves it from xx to jj.

The controller can pick any of ll internal slots, so there are ll alternative paths.

PB,TST=[1−(1−p′)2]l,p′=pclP_{B,TST} = \left[1 - (1 - p')^2\right]^{l}, \quad p' = \frac{pc}{l}

Three-stage STS network

 HWa->[S: a->m in i]->[TSI m: i->j]->[S: m->b in j]->HWb
  1. Input space stage sends the sample in slot ii to any free centre TSI mm.
  2. TSI mm changes the slot from ii to jj.
  3. Output space stage connects TSI mm to highway bb in slot jj.

There are kk alternative paths (centre TSIs):

PB,STS=[1−(1−p′)2]k,p′=pNkP_{B,STS} = \left[1 - (1 - p')^2\right]^{k}, \quad p' = \frac{pN}{k}

Comparison (example)

For p′=0.3p' = 0.3: a two-stage network blocks with probability about 0.3 (one link busy). A three-stage network with 8 paths gives (1−0.72)8=0.518≈4.6×10−3(1 - 0.7^2)^8 = 0.51^8 \approx 4.6 \times 10^{-3}, and with 32 paths about 4.4×10−104.4 \times 10^{-10}.

NetworkPathsBlocking
TS / ST1High (≈ pp)
STSkkLow, non-blocking if k≥2N−1k \ge 2N-1
TSTllVery low, non-blocking if l≥2c−1l \ge 2c-1

So 3-stage STS or TST networks remove the single-path problem of 2-stage networks; TST is preferred because time expansion is cheaper than space expansion.

✓ Answer checked by Claude Opus 5.5 (high effort)

All Telecommunications chapter 4 questions →