The invention relates to scheduling techniques for uplink transmission from a mobile device, wherein the mobile device uses a priority service via a network access point. As used herein, a priority service means a service to the packets of which some priority-based scheduling is applied.
Prior QoS techniques, one of which is disclosed in U.S. Pat. No. 6,738,361 to Jukka Immonen et al., are based on the assumption that a mobile subscriber determines a specific QoS for each service or type of services, and the subscriber is invoiced according to the determined QoS parameters. The assignee of this application has tested such QoS techniques, and surveys carried out among pilot subscribers have revealed that the pilot subscribers find such QoS techniques difficult to understand. Keeping track of a myriad of different subscriptions is a burden to access network operators.
An object of the present invention is to provide technique for implementing the method so as to alleviate the above disadvantages. In other words, the invention seeks to provide a scheduling technique that is more easily manageable to the subscribers and the operators.
The object of the invention is achieved by a method and software as disclosed in the attached independent claims. The dependent claims disclose specific embodiments of the invention.
The invention is partially based on finding a hidden problem, ie, the fact that the well-known “you get what you pay for” metaphor leads to very complex invoicing schemes. The invention is also based on the realization that service providers instead of the mobile subscribers determine appropriate QoS parameters for each priority service.
The invention is based on the idea that prior to a mobile device's access of a priority service, the access point serving the mobile devices polls the mobile device for scheduling information, ie, sends a request for scheduling information. The mobile device sends the access point the requested scheduling information. The access point uses the requested scheduling information to determine if priority scheduling is needed by checking the information against available packet filters for priority services. Next, the access point sends a scheduling decision to the mobile device, and the mobile device sends the access point one or more uplink data packets for accessing the priority service, wherein the uplink data packets are as specified in the scheduling decision by the access point. The headers of the received uplink data packets indicate the need for the priority service, which is verified by the access point to ensure that the priority scheduling is only used when appropriate.
The polling for scheduling information and filter configuration phases may be omitted if the network load remains below a determined threshold. For instance, if the network load is low enough that the network can process each packet according to the quality needed for the priority services, there is no need to perform extra steps to prioritize some packets over others. Omitting the polling phase saves capacity of the radio link and the battery of the mobile device.
Scheduling is only practical in a busy cell, where there are multiple mobile devices willing to send data packets. The scheduling decision determines in which order and how much radio capacity is allocated to each mobile device. This decision is based on the scheduling information matching the filter(s). Thus, if a mobile device has data packets to send that relate to a priority service, the scheduler will know this by the scheduling information extracted from the packet(s) matching a filter for a priority service. Based on this information the scheduler schedules, ie, assigns the radio resources, such that the priority service will get the capacity it needs in proper time. For example, packets with real-time requirements are prioritized over delay-tolerant packets.
The scheduling technique of the present invention is easier to manage than the prior art techniques because the number of priority services is probably smaller than the number of subscribers, which is why the invention requires fewer service-specific filters than does a technique in which the service-specific filters are associated with the subscribers.
An embodiment of the invention comprises configuring, in the access point, a predetermined filter for each of several priority services.
In an embodiment the access point may discard an uplink data packet or lower its priority if it does not correspond to the selected filter.
The access point may send the request for scheduling information in a point-to-point message to each individual mobile device, or to multiple mobile devices in a broadcast or multicast message. In one embodiment, such a message can be a part of the frame structure of the underlying radio link, or a part of a radio beacon or system information message, see document 3G TS 25.304, version 3.2.0, release 1999, for example.
The requested scheduling information may comprise any information about the packet or the packet header, including the amount of data to be sent by the mobile device and/or protocol type information and/or address information for the data packets. For example, the address information may comprise IP source or destination address.
The access point may encode the scheduling decision to the DSCP (Differentiated Services Code Point) field, by overwriting it with the value associated with the filter before forwarding the uplink data packet.
An aspect of the invention is a method or software to be executed by an access point. Another aspect is a method or software to be executed by a mobile device.
In the following the invention will be described in greater detail by means of specific embodiments with reference to the attached drawings, in which
In step 2-6, the access point 13 determines an appropriate filter based on the scheduling information sent by the mobile device 14. Let us first assume that a filter configuration substantially as shown in
If a filter configuration as shown in
In step 2-7, the access point AP sends the mobile device MD a scheduling decision, granting the mobile device MD some usage of the radio resources.
In step 2-8, the mobile device 2-8 sends data packets to the access point 13, in order to access the priority service at the server 11. In step 2-10 the access point 13 examines one or more parameters in the headers of the data packets (or one of the data packets) sent by the mobile device, in order to verify that the data packets sent in step 2-8 correspond to the scheduling information sent by the mobile device in step 2-4. If the data packets sent in step 2-8 correspond to the scheduling information sent earlier, the access point 13 processes the data packets according to the filter that was determined in step 2-6. This processing is denoted by reference numeral 2-10. For instance, such processing of packets may comprise overriding some priority-related parameters in the packet headers, as shown in more detail in
In one embodiment of the invention, the polling for scheduling information prior to actual service usage is only performed when the network load exceeds some predetermined threshold. As long as the network load remains below that threshold, steps 2-2, 2-4, 2-6 and 2-10 may be omitted. In one embodiment of the invention, also the step 2-7 may be omitted. This corresponds to the situation where there is no scheduling being performed by the AP at all, but MDs can send when they sense the frequency to be free (as in Ethernet). If the step 2-7 is NOT omitted, while the earlier steps are being omitted, the AP will schedule without any specific scheduling info, maybe in round-robin fashion, giving each terminal some resource that they then either use or not.
In
The access point 13 examines the header 322 of the mobile-originated data packet 32. In this embodiment, the access point 13 examines particularly the address field 323 and verifies that the contents of the address field 323 corresponds to the contents of the address field 313 of the filter 31. If the address fields 323 and 313 match, the access point 13 processes the data packet according to the filter. The address matching may be decided on the basis of either all or part of the bits in the two addresses having the same value, which covers the case wherein the filter specifies a wildcard instead of a full address. For example, a filter with an address of “123.24.*.*” would match any packets having an address which begins with “123.24”. In this embodiment, such processing comprises overriding the DSCP field 324 of the mobile-originated packet 32 with the DSCP′ field 314 of the filter 311.
The particular fields shown in
As shown in
In addition to the protocol changes shown in