1 Downlink Wireless Spectral Efficiency Normalized to WCDMA Downlink
10x
9x •Results obtained using different simulation 8x methodology are not directly comparable.
7x
6x •CDMA spectral efficiency would be much
5x higher using NGMN assumptions compared
4x to the results from 3GPP2 assumptions.
3x
2x
1x
0x 1x-Rel0 1x1 DOr0 1x1 DOrA 1x2 DOrB 1x2 Simulation 10x Methodology 9x .NGMN 8x
7x
6x .3GPP
5x
4x .3GPP2
3x
2x
1x
0x GPRS EDGE WCDMA WCDMA HSDPA HSPA HSPA+ HSPA+ LTE 1x2 LTE 2x2 1x1 1x2 1x1 Rel6 1x2 Rel7 1x2 Rel7 2x2
2 •2 x 10MHz (20MHz total) Spectrum assumed for all technologies Uplink Wireless Spectral Efficiency Normalized to WCDMA Uplink
6x •Results obtained using different simulation 5x methodology are not directly comparable.
4x •CDMA spectral efficiency would be much 3x higher using NGMN assumptions compared to the results from 3GPP2 assumptions. 2x
1x
0x 1x-Rel0 1x2 DOr0 1x2 DOrA 1x2 DOrA/B Full IC 1x2
6x Simulation Methodology 5x
4x .NGMN
3x .3GPP
2x .3GPP2 1x
0x GPRS EDGE WCDMA 1x2 HSPA Rel6 HSPA Rel6 LTE 1x2 1x2 Full IC 1x2
3 •2 x 10MHz (20MHz total) Spectrum assumed for all technologies Simulation Assumptions (1)
. Comparing results across technologies needs to take simulation assumptions into considerations . UMTS/HSPA(+) and LTE results are based on NGMN or 3GPP evaluation methodology . CDMA 1X and EVDO simulations uses 3GPP2 evaluation methodology . CDMA spectral efficiency would be much higher using NGMN methodology compared to the results from 3GPP2 assumptions
High level difference of evaluation methodologies
3GPP2 NGMN 3GPP Cell site-to-site 2.0 0.5 * 2.8* distance (km) Channel model Mix of Ped, Veh One channel model: Mix of Ped and and Rician Urban Macro 3km/h Vehicle models models * Values shown are used to generate previous spectral efficiency results. Other site-to-site distance can be used in 3GPP and NGMN methodology 4 Simulation Assumptions (2)
. A total of 2 x 10MHz (20MHz total) spectrum allocation is assumed for all technologies . 7 EV-DO carriers . 2 WCDMA carriers . 1 LTE carrier (2 x 10MHz) . The spectral efficiency results are obtained in interference-limited scenarios typically seen in urban deployments . Rural deployments may not be interference limited . Traffic type is assumed to be full-buffer . Spectral efficiency results will be changed for different type of traffic, e.g. bursty traffic . Spectrum efficiency also depends on number of antennas and receiver type . UL results of HSPA rel 7 and DO rev A/B with full IC assume the BS receiver performing full interference cancellation of pilot, overhead and traffic channels . The number of receiver antennas of BS and mobile is limited to 2 or less, which is typical in current network
5