All answers have been held as short as possible and require an understanding and study of the corresponding chapter of the book.
The GSM network was mainly designed for narrowband circuit switched communication. UMTS on the other hand has been designed from the beginning not only for voice communication but also for packet switched data transmission at high speeds. To achieve this goal, UMTS uses an air interface technology known as code division multiple access (CDMA). CDMA does not assign a specific frequency and timeslot to a single user for transferring information like in GSM. Instead a unique channelization code is assigned to each user. In a UMTS network, all users transfer their data simultaneously and the network is able to filter each data stream out of the result by applying the different codes on the received signal.
The use of CDMA, which has been described briefly in answer 1, in combination with codes and variable code lengths has a number of advantages. One of those are the improved data rates compared to GPRS as well as shorter round trip delay times. In a first step, speeds up to 384 kbit/s can be achieved per user in downlink direction. This enables a number of new applications which require a broadband Internet connection which can now be used while on the move, like for example in cars and trains. Furthermore, UMTS also supports 64 kbit/s circuit switched channels on the air interface in both uplink and downlink direction which are used for video telephony.
With a R99 access network, data rates of up to 384 kbit/s in downlink and 128 kbit/s in uplink (most network allow only 64 kbit/s in uplink direction) are possible. In practice, however, such channels are no longer used today.
OVSF is the abbreviation for orthogonal variable spreading factor. OVSF allows to assign different code lengths to different users depending on their application. A long spreading factor is used for applications such as voice telephony which require only a small amount of bandwidth. Shorter codes are used for faster data transfers with speeds of 128 kbit/s or 384 kbit/s.
Scrambling in downlink direction is necessary to allow mobile devices to distinguish different cells of a network which send on the same frequency. Without scrambling codes it would not be possible to use the complete code tree in each cell. In uplink direction, a scrambling code is required in order to avoid interference problems generated by users which are at different distances to a base station. By using scrambling codes, orthogonality in uplink direction is preserved. Scrambling is also required as not all channelization codes generate a pseudo random bit pattern which is required for spectral distribution, as many consecutive bits can have the same value.
During times of low activity a cell can cover a wide area as interference is low and thus distant devices can receive data without problems. During times of high activity distant devices are not able to receive data correctly anymore due to the interference caused by devices closer to the base station. Thus, the area covered by the cell is smaller than before. As the cells coverage area shrinks during high traffic loads, this phenomenon is also called cell breathing.
While in Cell-DCH state, a dedicated channel with an individual code is assigned to a mobile device. Data can be sent to and from the mobile device without prior reservation of resources. In Cell-FACH state on the other hand the mobile device sends and receives frames via the RACH and FACH. These channels are shared between several devices. Thus, no dedicated code is assigned in this state to a user which implies that bandwidth and round trip delay times can not be ensured. Furthermore, the available bandwidth is quite limited.
In PMM connected state a device can either be in Cell-DCH, Cell-FACH, Cell-PCH or URA-PCH state.
While in soft handover state a mobile device communicates with several cells simultaneously. A handover can thus be performed without any interruption of the ongoing transmission. The mobile device only uses the minimal power required to remain connected to one of the cells taking part in the soft handover. When transmission conditions change, a different cell can take over the connection very quickly. A disadvantage of the soft handover is the use of additional resources on the air interface. If too many cells are part of a soft handover procedure for a single mobile device, the available capacity for other subscribers can be severely reduced.
An SRNS relocation is performed when all current cells of a subscriber are controlled by a drift RNC. This can happen for example if a subscriber is moving far away from the location at which the connection was initially established. By performing an SRNS relocation one of the drift RNCs becomes the new serving RNC. This means that the routing between the MSC and the SGSN on the one side and the old and new RNC on the other side has to be changed.
In Cell-FACH state, mobility management is not performed by the network but by the mobile device itself. If the mobile device detects that a different cell would give a better service, it autonomously performs a cell change. Once the cell change is performed the mobile device reports to the network and all data is afterwards exchanged via the new cell.
As a mobile device in Cell-DCH transmits data continuously, it has no possibility to search for other cells on other UMTS or GSM frequencies. If a subscriber moves to the border of the UMTS coverage area or if UMTS cells in the subscribers area use a different frequency, the network can instruct the terminal to activate compressed mode. While in compressed mode, silence periods are inserted in certain intervals to allow the mobile device to search for neighboring cells on different frequencies. The measurement results generated during those periods are then sent to the network which can then use them for the decision making on when and how to perform a handover to one of these cells.
HSDPA uses adaptive modulation and coding to quickly react to changing conditions on the air interface. Thus, less error correction and detection bits can be used while the radio link is stable which increases speed under such conditions. Instead of only using one data stream, HSDPA users can receive data via several code channels simultaneously (multi code). Additionally, Hybrid Automatic Retransmission Requests (HARQ) are used to detect transmission errors very quickly and to retransmit the data before higher layers detect a problem. Finally, intelligent scheduling can be used to reduce the data rate to a subscriber while signal conditions are temporarily bad and to increase the data rate again once conditions have improved. This improves overall cell capacity and in turn also increases the average date rate available to all users in a cell.
HSDPA supports a simultaneous dedicated channel for voice calls along one or more HSDPA channels. Thus, a subscriber can use both services at the same time.
Without an E-DCH, uplink packet speeds are limited to 64 kbit/s -128 kbit/s in most networks and 384 kbit/s in rare cases. The E-DCH concept increases the uplink speed per user to up to 2 Mbit/s in ideal conditions and to around 800 kbit/s und normal conditions in operational networks. Additionally, E-DCH also increases overall system capacity in the uplink, so more users can access the network with higher uplink speeds at the same time. This is very important for applications that send as much data as they receive. Such services are for example voice and video telephony.
In uplink direction, the Node-B schedules different subscribers by assigning a certain transmission power level to each subscriber. The mobile station then has to select a corresponding code length and coding scheme. Thus, the transfer speed and implicitly the noise that is generated by the subscriber for other terminals in the network is limited. Power assignments can be given as absolute grants or as relative grants, which increase or decrease the power level for a mobile station. Neighboring cells which are part of the active set can also send relative grants to decrease the power level if a mobile station produces too much interference for subscribers in its area.