Page 288 - From GMS to LTE
P. 288

274  From GSM to LTE-Advanced Pro and 5G

            The MME then terminates the attach process with a reject cause 15 (no suitable cells in
            this tracking area) that triggers the mobile device to change to UMTS or GSM and to
            perform a new attach procedure there.
             Once authentication has been performed, a set of session keys are generated, as
            described in more detail in [20]. Afterward, ciphering and integrity protection can be
            activated for all NAS messages between the UE and the MME. While integrity checking
            is mandatory, the use of ciphering for signaling messages between the mobile device
            and the MME is optional. Once the corresponding keys are known by the eNode‐B, it
            will also activate integrity checking and ciphering for RRC messages and ciphering for
            the user data bearer over the air interface. As NAS messages are carried inside RRC
            messages, they are ciphered twice if encryption for signaling messages was activated in
            the previous MME/UE security exchange. In any case, two integrity checks are per-
            formed, one between the UE and the eNode‐B, and another one between the UE and
            the MME.
             When ciphering and integrity checking are activated, the UE, MME and eNode‐B can
            select an appropriate EPS Encryption Algorithm (eea0, eea1, eea2, etc.) and an EPS
            Integrity Algorithm (eia1, eia2, etc.) from a list of algorithms that are supported by both
            sides. Eea0 corresponds to no encryption being used. Therefore, in operational
              networks, the use of eea0 between the mobile device and the eNode‐B should be the
            exception. Integrity checking is always used even if encryption is not activated. This is
            why eia0 does not exist. Eea1/eia1 corresponds to the algorithms introduced in 3GPP
            Release 7 for UMTS (UEA2, SNOW3G).


            4.9   Interconnection with UMTS and GSM


            When a mobile device is at the border of the coverage area of the LTE network, it should
            switch to another network layer such as UMTS and GSM to maintain connectivity. In
            the worst case, the mobile device loses LTE network coverage and if it does not find a
            suitable LTE cell on the current channel it will search for LTE cells on other channels
            and also switch to other frequency bands and other RATs to regain contact with the
            network. This would take a significant amount of time, typically between 10 and 30
            seconds if no information on alternative cells was previously received from the network.
            During this time the device is not reachable for services trying to contact it, such as
            push e‐mail or incoming voice calls. It is therefore better if the network supports the
            mobile device in finding other suitable channels, bands or radio technologies. There are
            three basic procedures for these purposes, which are described in the following section:
               cell reselection from LTE to UMTS or GSM;
            ●
               RRC connection release with redirect from LTE to UMTS or GSM;
            ●
               inter‐RAT handover from LTE to UMTS.
            ●
             Irrespective of whether the mobile has to find a GSM or UMTS network by itself or is
            supported by the network, the LTE network has to be connected with the GSM and
            UMTS networks so that the subscriber’s context, that is, the assigned IP address, QoS
            settings, authentication and ciphering keys, and so on can be seamlessly exchanged
            between all core network components involved. In practice core networks use  combined
            GSM, UMTS and LTE network nodes today and hence, the interfaces described in
   283   284   285   286   287   288   289   290   291   292   293