TD-SCDMA based SFN network and associated methods
Networking methods of a Single Frequency Network (SFN) in communication systems are disclosed.
1. A method for a single frequency network (SFN) based on a TD-SCDMA network, the method comprising:
dividing the TD-SCDMA network into two layers: one layer for a mobile communication network, and the other layer for a TD-SCDMA single frequency network (SFN), wherein the TD-SCDMA SFN is independent from the mobile communication network in which some frequency resources are reserved for the SFN only, or a partial timeslot resources are used for the mobile communication network, and the rest of the timeslot is used for the SFN service;
performing channel estimation; and
thereafter, performing signal detection using an equalization algorithm, wherein the equalization detection algorithm includes time equalizing and frequency equalizing, wherein performing signal detection further comprising:
sending received data to a data splitting unit where user data and midamble data are separated;
sending the midamble data to a channel estimation unit for channel response estimation;
sending the user data together with the estimated channel to a channel equalizer unit for equalization;
sending output data from the channel equalizer to a de-scrambling unit for chip-level de-scrambling;
sending chip de-scrambled data to a de-spreading unit for de-spreading;
soft de-modulating de-spread data in a soft demodulation unit to obtain soft bits; and
further processing the soft bits for decoding and service recovering.
2. The method of claim 1 , wherein the estimated channel with total time duration of L max +16 chips is obtained from the channel estimation unit; the estimated channel h is noise-depressed; 64 zeros are inserted to the noise-depressed channel h, after which FFT is performed as: H=FFT(h(1),h(2), . . . , h(L max +16),0,0, . . . ,0); then IFFT is performed to H as:
w
=
I
F
F
T
(
1
H
)
;
the equalizer is formed by w as a FIR filter structure.
3. The method of claim 2 , further comprising:
calculating channel taps power P 1 P 2 . . . P L max +16 from the estimated channel h(k) with L max +16 chips time duration;
determining a maximum power as P=max(P 1 P 2 . . . P L max +16 );
forming a noise-depressing threshold as: τ=γP , where γ is a scalar value belonging to (0 1); and
setting an index as I={i|conditioned on P 1 <τ,i=1,2, . . . ,L max +16}, and setting weak taps of the channel to zeros as h(i)=0,iεI.
4. The method of claim 1 , wherein the estimated channel with total time duration of L max +16 chips is obtained from the channel estimation unit; the estimated channel h is noise-depressed; Z transform is performed as: H(z)=Z(h(n)); then the equalizer is formed by
1
H
(
z
)
as an IIR filter structure.
5. The method of claim 1 , wherein the estimated channel with total time duration of L max +16 chips is got from the so-called channel estimation unit; the estimated channel h is noise-depressed; Z transform is performed as: H(z)=Z(h(n)); then the equalizer is formed by
H
*
(
z
-
1
)
H
*
(
z
-
1
)
*
H
(
z
)
+
δ
n
2
(
z
)
as a FIR/IIR hybrid filter structure, where δ n 2 (z) is the z transform of additive white noise.
6. A method for a single frequency network (SFN) based on a TD-SCDMA network, the method comprising:
dividing the TD-SCDMA network into two layers: one layer for a mobile communication network, and the other layer for a TD-SCDMA single frequency network (SFN), wherein the TD-SCDMA SFN is independent from the mobile communication network in which some frequency resources are reserved for the SFN only, or a partial timeslot resources are used for the mobile communication network, and the rest of the timeslot is used for the SFN service;
performing channel estimation; and
thereafter, performing signal detection using an equalization algorithm, wherein the equalization detection algorithm includes time equalizing and frequency equalizing, wherein an equalization algorithm of OFDM is applied for receiving and demodulating the TD-SCDMA SFN receiver; when parameter L max >0, a delay of an estimated channel multi-path is larger than 16 chips, received signals after a start of midamble data with L max +16 chips delay to the next timeslot's midamble data forms an OFDM symbol.
7. The method of claim 6 , wherein when L max >0, received signals of a TD-SCDMA downlink transmission is regarded as an OFDM symbol with a length of 864 chips or 1024 chips.
8. The method of claim 7 , wherein the OFDM symbol with a length of 864 chips comprises: the last part of the midamble data with length of 128-L max chips, a second user data part of a first timeslot with length of 352 chips, a guard period part with length of 16 chips, a first user data part of a second timeslot with length of 352 chips, and the first 128-L max chips of a midamble data part of the second timeslot.
9. The method of claim 7 , wherein the OFDM symbol with length of 1024 chips comprises: the last 128-L max chips of TS 0 midamble data, second part user data of TS 0 , the guard period of 16 chips, a DWPTS with length of 96 chips, the guard period of 32 chips, inserted zeros with length 32 chips, a first part of user data of TS 1 , and first L max +16 chips midamble data.
10. The method of claim 6 , wherein a receiving method of the SFN broadcast based on TD-SCDMA network, is characterized in that: when L max ≦0, received signals from TD-SCDMA downlink transmission in one downlink timeslot is regarded as an OFDM symbol with a length of 864 chips.