A mobile station (100) and a compression/decompression node in a wireless network (5), preferably the GGSN 30 (30), agree on whether to apply compression/decompression, and what compression/decompression scheme to use either by pre-configuration (when the GGSN 30 is in the home network of the mobile terminal) or by signalling (when the GGSN 30 is in a visited or a roamed-to network). Additional information, such as a profile-specific dictionary, may be used to increase the compression efficiency, particularly during start-up. Dynamic data can also be accommodated. The compression/decompression may be turned on and off on a packet-by-packet basis, depending on whether the compression actually results in a smaller packet. A compressed/uncompressed indicator bit or bits can be inserted explicitly on a per packet basis, and/or the compressed/uncompressed indication can be derived implicitly. A desired data compression/decompression algorithm, or any desired Proprietary Feature, can be signaled, possible without requiring standardization changes, by using a Packet Data Protocol PDP Context Activation message with a predetermined Access Point Name (APN) message information element that is sent from a first network element to a second network element for at least identifying the desired compression/decompression algorithm. It another embodiment, for the case where Proprietary Feature capability is exchanged between a RAN and a UE network element, a transmission of an extension to a UE Capability Enquiry message is sent to the UE for identifying the Proprietary Feature.