Frequency/Power Profile

As with IMT2000DS, CDMA2000 is tightly specified in terms of spurious emissions,
measured both for their impact in-band and out-of-channel, as shown in Figure 3.29.
In markets with legacy 30 kHz channel-spaced networks (US TDMA 800 MHz and
1900 MHz), adjacent channel power ratios need to be qualified with respect to adjacent
narrowband channels. Similar specifications are required for out-of-band performance.
The CDMA2000 specification requires spurious emissions outside the allocated
system band (measured in a 30 kHz bandwidth) to be 60 dB below the mean output
power in the channel bandwidth or -13 dBm, whichever is smaller.
Frame erasure rate can be used as a measure of receiver performance, provided coding
and error correction is applied equally to all bits�"that is, there are not classes of
bits with different levels of error correction. Frame erasure rate is the ratio of the number
of frames of data received that are deleted because of an unacceptable number of
errors to the total number of frames transmitted. Frame erasure rate is used as a measure
of receiver performance.

We can use frame erasure rate to measure sensitivity and dynamic range, spurious
immunity, and performance in AWGN and fading channels. CDMA2000 uses 20 ms
frames. Base station receiver performance is expressed in terms of FER versus Eb /No.
The Eb /No required will be a function of data rate and channel requirements.
At system level, the use of a continuous pilot in CDMA2000 provides better channel
sounding, compared to IMT2000DS, but it uses more transmit energy. The continuous
common pilot channel provides the following:
 More accurate estimation of the fading channel
 Faster detection of weak multipath rays than the per-user pilot approach
 Less overhead per user
Turbo coding is used for higher data rates with K = 9 constraint length.
The forward link coding is adaptive. Interleaving can either be 20 or 5 ms. A6-bit, 8-
bit, 10-bit, 12-bit, or 16-bit CRC is used for frame error checking with 1/2, 1/3, 1/4 rate
K=9 convolutional coding. Equivalent rate turbo codes are used on supplemental
channels. Each supplemental channel may use a different encoding scheme. Similarly,
downlink coding is adaptive, using a 6-bit, 8-bit, 10-bit, 12-bit, or 16-bit CRC for frame
error checking, and 9/16, 1/2, 1/3, 1/4 rate K = 9 convolutional coding. Equivalent
rate turbo codes are used on supplemental channels. Each supplemental channel may
use a different encoding scheme. Interleaving is again either 5 ms or 20 ms.
Closed-loop power control is carried out at an 800 Hz control rate. The open loop sets
Tx power level based on the Rx power received by the mobile and compensates for path
loss and slow fading. The closed loop is for medium to fast fading and provides compensation
for open-loop power control inaccuracies. The outer loop is implementationspecific
and adjusts the closed-loop control threshold in the base station to maintain the
desired frame error rate. The step size is adaptive, either 1 dB, 0.5 dB, or 0.25 dB. As with
IMT2000DS, power control errors will directly subtract from the link budget.
The power control dynamic range is as follows:
 Open loop ±40 dB
 Closed loop ±24 dB
Power control errors are typically 1.3 dB (low mobility) or 2.7 dB (high mobility).
Dynamic range is similar to other existing networks:
Mobile 79 dB
Base station 52 dB
FDD isolation (45 MHz, 800 MHz, 80 MHz at 1900 MHz)
Class II mobile 55 dB Tx to Rx
Base 90 dB (higher effective power, 5 dB lower noise floor)
Class IV handsets are equivalent to Power Class 3 handsets in IMT2000DS (250
mW). Class V handsets are equivalent to Power Class 4 handsets in IMT2000DS (125
mW). Both networks also support higher-power mobiles.
Class I: 28 dBm < EIRP < 33 dBm (2 W)
Class II: 23 dBm < EIRP < 30 dBm
Class III: 18 dBm < EIRP < 27 dBm
Class IV: 13 dBm < EIRP < 24 dBm (250 mW)
Class V: 8 dBm < EIRP <21 dBm (125 mW)
CDMA 1xEV has a high data rate option for the downlink, separate 1.25 MHz RF
channel, QPSK, 8 PSK, 16-level QAM, and evolution to meet IMT2000MC requirements
(3xRTT). 1xEV adds adaptive modulation as a mechanism for increasing data
throughput.

The Media Access Control (MAC) layer in IS2000 manages code allocation (the provision
of physical layer resources to meet application layer requirements). An active
high-rate mobile assigned a fundamental channel on origination negotiates high data
rate service parameters. The mobile then sleeps but remains locked to a low-rate channel
for synchronization and power control.
The handset signals a high data burst request by indicating to the base station (BS)
its data backlog and maximum data rate requested. The handset includes pilot
strength information for cells in its neighbor list, which indicates local interference levels.
Additionally pilot strength measurements allow the base station to qualify instantaneous
downlink capacity.

Supplemental code channels can then be allocated as required. In Figure 3.30, the
handset communicates on the fundamental code channel with two base stations (BS1
and BS2). During a burst transmission, one or more supplemental code channels are
assigned at BS1, BS2, or both. The MSC performs distribution on the forward link and
selection on the reverse link. The Radio Link Protocol (RLP) does an Automatic Repeat
Request (ARQ) and the interworking function (IWF) provides access to the packet data
network.
When there is backlogged data, the mobile goes into active mode. If backlogged data
exceeds a threshold, the mobile requests a supplementary channel (SCRM), sent on the
fundamental code channel. The BS/MSC uses pilot strength measurements made by
the mobile to decide on burst admission control and allocates supplementary channels.
When backlogged data at the IWF exceeds a predetermined threshold, the IWF initiates
a request for supplementary channels. The mobile is paged if not already in an
active state.
In IS95B, a mobile is either active or dormant, and in CDMA2000, a handset can go
into control hold, maintaining a dedicated control channel and power control (burst
transmission with no added latency). In suspended state, there are no dedicated channels,
although a virtual set of channels are maintained. In dormant state, there are no
pre-allocated resources; in other words, the deeper the sleep, the lower the power consumption,
but the longer it takes to wake up. 111