Privacy-Preserving in Mobile Crowdsensing
Blockchain-Based Decentralized and Lightweight Anonymous Authentication for Federated Learning
ShadowEth: Private Smart Contract on Public Blockchain
Nereus: Anonymous and Secure Ride-Hailing Service based on Private Smart Contracts
POSE: Practical Off-chain Smart Contract Execution
Ekiden: A Platform for Confidentiality-Preserving, Trustworthy, and Performant Smart Contracts
https://www.sigsac.org/ccs/CCS2023/
https://www.usenix.org/conference/usenixsecurity24
CFP: https://www.usenix.org/sites/default/files/sec24_cfp_101123.pdf
只收集English Language Proficiency要求和DDL
12.1 (来年9月)
6.1 (来年1月)
10.1 (来年5月)
2.1 (9月)
6.1 (来年1月)
10.1 (来年5月)
2.1 (9月)
6.1 (来年1月)
10.1 (来年5月)
2023.9.15 - 2023.12.15 (2024.9)
2023.9.15 - 2023.12.15 (2025.1)
Application open dates and deadlines for an upcoming intake have not yet been configured in the admissions system. Please check back later.
1.10 (9月)
6.1 (1月)
1.31 (9月)
6.1 (1月)
10.1 (5月)
2022.12.1 (Fall 2023)
No mentioned. TOEFL 93 (Wrting 22, Speaking 22) link
2023.1.5 (Fall 2023 & Winer 2024)
TOEFL 90 (每项最少20)
January – apply by June 15
May – apply by October 15
September – apply by December 15
TOEFL 90 (每项最少20)
September entry – apply by June 1
January entry – apply by October 1
May entry – apply by February 1
MMU是某个车辆通过DSRC通信后，得到周围可用车辆的list，然后维护所有可用的booth，为共识服务提供接口。
V-Guard develops a Membership Management Unit (MMU) that keeps track of available vehicle connections and manages membership profiles. The MMU describes a membership profile that contains a set of vehicles as a booth. Below illustrates the management of booths (of size 4) when vehicles are communicating via DSRC.
V2X有两种通信技术：DSRC和C-V2X
DSRC和C-V2X的区别是：
标准不同：DSRC的通信标准由美国交通部制定的，是IEEE 802.11p；而C-V2X的技术标准由3GPP制定的，3GPP Rel. 14和15。
频段不同：DSRC的频段是5.9GHz，而C-V2X使用LTE或5G的移动通信频段。
通信模式不同：DSRC采用点对点（P2P）和点对多点（P2MP）通信模式，而C-V2X采用蜂窝网络通信模式。
技术架构不同：DSRC需要单独的通信设备和基础设施来实现通信，而C-V2X则可直接利用现有的蜂窝网络基础设施进行通信，无需额外的设备。
支持的应用不同：DSRC主要用于车辆安全应用（如协同自适应巡航、自动紧急制动等），而C-V2X还支持更广泛的车联网应用，例如交通管理、自动驾驶等。
技术能力不同：C-V2X相比DSRC有更高的数据传输速率和更低的延迟，可以支持更多的车辆和更大范围的通信。同时，C-V2X还能够与4G、5G等其他通信技术相互协作，为车联网提供更广泛的应用。
DSRC VS C-V2X: https://auto-talks.com/technology/dsrc-vs-c-v2x/
V2X-DSRC介绍https://zhuanlan.zhihu.com/p/541727482
DSRC是什么 及 在技术体系中的定位https://blog.csdn.net/m0_37495408/article/details/104823812
Dedicated short-range communications (DSRC) is a technology for direct wireless exchange of vehicle-to-everything (V2X) and other intelligent transportation systems (ITS) data between vehicles, other road users (pedestrians, cyclists, etc.), and roadside infrastructure (traffic signals, electronic message signs, etc.).[1] DSRC, which can be used for both one- and two-way data exchanges, uses channels in the licensed 5.9 GHz band. DSRC is based on IEEE 802.11p. (Wiki-Dedicated short-range communications
Github: eigenlayer-contracts/docs/EigenLayer-tech-spec.md
官方doc: https://docs.eigenlayer.xyz/overview/readme
[登链] 2023/04/28期：深度解析以太坊质押和LSD（文字整理版）: https://learnblockchain.cn/article/5740
[登链] 深入了解再质押（Restaking）：EigenLayer https://learnblockchain.cn/article/5732
详解 EigenLayer：通过“再质押”扩展以太坊信任层: https://new.qq.com/rain/a/20230404A02TU700
EigenLayer is a set of smart contracts on Ethereum that allows consensus layer Ether (ETH) stakers to opt in to validating new software modules built on top of the Ethereum ecosystem.
Ethereum 提出了新的信任模组化创新，使得开发DApps 不再需要重新打造一个去中心化网路。可以直接使用在Ethereum 网路所提供的信任。
Ethereum 只提供区块生成层面的信任，Middlewares 无法利用Ethereum 的信任机制来打造更广泛的去中心化服务。Middlewares 仍需另外打造自己的信任网路（AVS），这中间的主要成本来自维持网路安全的经济模型。
EigenLayer 是一个平台，旨在探讨Liquid Staking Derivatives 板块，并提供更安全、更可靠的去中心化服务。其核心概念是Restaking，即让同一份资金可以质押在不同的信任网路，来担保不同信任网路的安全性。
Pooled Security via Restaking 可以让质押者获得更多收益，同时也降低了维持DApp 安全性的成本。透过Restaking，质押者可以用同一份资金质押到不同的信任网路，担任不同的网路的验证者来获得验证奖励。而对于中介服务（Middlewares）来说，Restaking 也能够让建立信任网路的门槛大幅降低，使其能够继承Ethereum 网路部分或全部的信任（资金）。
Free-market governance 则建立一个开放市场的自由机制。透过这种机制，Stakers 能根据收益和风险来决定要质押的对象。Middlewares 能够根据自己的需求限制Staker 来保证自身网路的安全性。
EigenLayer 创新也带来一些额外的问题和风险，例如Operator collusion 和Unintended Slashing。面对这些问题我们仍需要根据不同的使用场景采用不同的手段来解决。
Reward:
在Ethereum上质押本身带来的收益
再质押到Dapp上带来的收益
(如果选择了代理，可能要考虑代理的分红问题，这个eigenlayer中不涉及设计，如果自己是自己的operator，那么应该没有这个问题)
Developers choose for themselves which tokens to accept as stake for their AVS. They may also choose whether there is preferential weighting for rewards they distribute to different types of staked tokens. (来自白皮书)
Punishment:
EigenLayer有Slashing机制，Slashing机制还有风险规避（诚实节点被slash）机制，但机制比较简单
As shown in Fig. 5b, EigenLayer provides multiple pathways for yield stacking which allow stakers to earn additional yield from securing new AVSs. Broadly we can think of three distinct layers of the blockchain: core protocol, AVS, and DeFi. Liquid staking can be thought of as stacking yield by first going to the core protocol and then the DeFi layer. Superfluid staking can be thought of as first going through the DeFi layer before going to the core protocol layer. In EigenLayer, there can be several modalities of restaking:
Solo Staking: In this model, solo stakers, who are natively restaked, have two options for participating in EigenLayer: (1) solo stakers can can opt in to AVSs on EigenLayer for which they can provide validation services directly; or (2) solo stakers can delegate EigenLayer operations to a different entity, while continuing to validate for Ethereum themselves. This latter option allows for home stakers with non-upgradeable / lightweight setups to continue contributing decentralization and censorship resistance to Ethereum (as well as not sharing the core Ethereum yield with any operator), while receiving additional rewards through EigenLayer via delegation to another operator. We envision that there will be many services built on EigenLayer that are designed to be lightweight, and suitable for home stakers who do not want to delegate their stake to another operator. For example, consider a decentralized price oracle, which is computationally easy to run, but requires high trust.
Delegation Model: The delegation model in EigenLayer requires that delegate their stake to an operator. If their operator doesn’t fulfill its obligations in the EigenLayer modules that it is participating in, then their deposited stake will be subject to slashing. Restakers who have delegated their stake with this operator will be slashed too. Hence, EigenLayer restakers should only delegate to trusted operators who have a track record of successfully fulfilling their obligations. There are no incentives built into EigenLayer for delegating, but is possible for others to build innovative delegation frameworks on top of EigenLayer.
Fee model: In addition, EigenLayer restakers need to take into consideration the ratio of fees that operators share back to delegators. Given that there will be many operators to which a restaker can delegate its ETH or LSTs, this will organically give rise to a free market of delegation between restakers and operators in EigenLayer. Each EigenLayer operator will put up a delegation contract on Ethereum that specifies how fees would be split back to delegators, and that delegation contract will route fees accordingly.
EigenLayer 中存在两类风险：
运营商串通
在现实中，只有一部分运营商选择加入给定的 AVS，其中一些运营商可能串通从一组 AVS 中窃取资金，然后会出现复杂的攻击。
意外罚没
在 AVS 及其相关基础设施和合约经过实际测试之前，许多罚没风险需要控制以避免产生更大的叠加风险。 一种风险是 AVS 创建时的意外罚没漏洞（例如，代码 bug），这种漏洞一旦触发会导致诚实运营商损失资金。
我们在这里提出两条解决方案：
Stakers
A staker is any party who has assets deposited into EigenLayer. In general, these could be any mix of ERC20 tokens and/or staked ETH itself (deposited by transferring withdrawal credentials to EigenLayer or depositing to the Beacon Chain through EigenLayer). Stakers can delegate their stake to an operator, or act as an operator themselves.
Operators
Operators in EigenLayer are those users who actually run the software built on top of EigenLayer. Operators register in EigenLayer, allowing stakers to delegate to them, and then opt-in to any mix of services built on top of EigenLayer; each service that an operator chooses to serve may impose its own slashing conditions on the operator.
Watchers
NOTE: at present, EigenLayer does not feature any optimistically rolled up claims. This paragraph reflects a potential future state of the system.
Some operations in EigenLayer are “optimistically rolled up”. This is a design pattern used where it is either impossible or infeasible to prove that some claim is true, but easy to check a counterexample that proves the claim is false. The general pattern is:
Watchers are parties who passively observe these “rolled up” claims, and step in only in the case of an invalid or false claim. In such a case, an honest watcher will perform the fraudproof, disproving the claim.
Services / Middleware
We refer to software built on top of EigenLayer as either services or middleware. Since we anticipate a wide variety of services built on top of EigenLayer, the EigenLayer team has endeavored to make a minimal amount of assumptions about the structure of services.
The StrategyManager contract keeps track of all stakers’ deposits, in the form of “shares” in the Strategy contracts. Stakers who wish to deposit ERC20 tokens can do so by calling the StrategyManager, which will transfer the depositor’s tokens to a user-specified Strategy contract, which in turn manages the tokens to generate rewards in the deposited token (or just passively holds them, if the depositor is risk-averse or if the token lacks good reward-generating opportunities).
Each Strategy contract is expected to manage a single, underlying ERC20 token, known as the underlyingToken.
Each user’s holdings in the strategy is expected to be reflected in a number of shares, and the strategy is expected to define methods for converting between an amount of underlying tokens and an amount of shares (and vice versa), somewhat similar to an ERC4626 Vault but without most of the tokenizing aspects of EIP-4626 (e.g. no transfer or transferFrom functions are expected). Assets may be depositable or withdrawable to a single Strategy contract in multiple forms, and the strategy may either actively or passively manage the funds. Since individual users’ share amounts are stored in the StrategyManager itself, it is generally expected that each strategy’s deposit and withdraw functions are restricted to only be callable by the StrategyManager itself.
Any staker in EigenLayer may choose to become either:
OR
(120-145w)
School Announcement:
In the reading, we know that the university has decided to … .
(First, it claims that ….; Second, it will ….)
Letter/Email:
The student/professor in the letter/email suggests that the university should(/n’t) … for 2 reasons.
First, … ; Second, ….
The woman/man in the conversation agrees/disagrees with the plan/letter for 2 reasons.
First, she states that … ; The student points out that … .
In addtion, she declares that … ; For example, she mentions … .
(Optional) That’s why she’s against/in favor of the plan/.
==================================================================================
(120-145w)
Well, in the XXX lecture, the professor used two examples to explain this concept.
The first example is … .
The second example given is … .
==================================================================================
(120-145w)
The professor in the XXX lecture talks about XXX from 2 aspects.
The first aspect/advantage/disadvantage/feature/benefit/reason is that … . For example, ….
While, the second XXX is that …. . For example, ….
(around 300 words)
In the reading material, the author cites three reasons to support the viewpoint that (summary of reading intro). Nevertheless, the lecturer in the listening material argues that (summary of listening intro), and opposes the points in the reading material.
First and foremost, the author argues that (reading point 1 + flow of idea). However, the lecturer offers an opposite stand that (listening point 1 + details).
In addition, the lecturer casts doubt on the author’s another reason that (reading point 2 + flow of idea), demonstrating that (listening point 2 + details).
Finally, the author’s third reason that (reading point 3 + flow of idea) is also counteracted by the lecturer who points out that (listening point 3 + details).
An apparent divide exists between the author and the lecturer on the evaluation of (content).
mooc模板对阅读部分内容更为精炼（时间较少），更多侧重于听力细节，难度稍高
huge版涉及更多阅读内容，顺序更为连贯，难度稍低
School Announcement:
Letter/Email:
The woman/man in the conversation agrees/disagrees with the plan/letter for 2 reasons. First, …. ; Second, … .
T3-Reading
In this set of material, the reading passage talks about (文章标题)… , which means (+定义句)
T3-Lecture
Well, in the XXX lecture, the professor used an example of … to explain this concept.
展开例子具体内容即可。
T4-Lecture
The professor in the XXX lecture talks about XXX from 2 aspects.
The first aspect/advantage/disadvantage/feature/benefit/reason is that … for example, ….
The second XXX is that .. for example, ….
In the reading material, the author cites three reasons to support the viewpoint that (summary of reading intro). Nevertheless, the lecturer in the listening material argues that (summary of listening intro), and opposes the points in the reading material.
First and foremost, the author argues that (reading point 1 + flow of idea). However, the lecturer offers an opposite stand that (listening point 1 + details).
In addition, the lecturer casts doubt on the author’s another reason that (reading point 2 + flow of idea), demonstrating that (listening point 2 + details).
Finally, the author’s third reason that (reading point 3 + flow of idea) is also counteracted by the lecturer who points out that (listening point 3 + details).
Both the reading and lecture are talking about 0, but the speaker disagrees with the perspective made in the reading because of … and …, which disagrees with the standpoint in the reading that ….
The reading first claims that 1. However, the instructor objects it with a new proof. The professor says that A because … That is why the opinion of the reading is challenged heavily.
Furthermore, the reading conceives that 2. In contrary, the professor asserts that B, which differs the reading as well. He believes that ….
Finally, the professor in the listening material, maintains that C since …, while the reading claims that 3. Therefore, the reading passage has been contradicted again.
Therefore, the speaker departs totally from the opinion on the topics made in the reading.
Both the reading and the lecture are talking about 0, and the speaker agrees with the perspective made in the reading.
The reading first claims that 1. Similarly, the instructor supports it with a piece of new evidence. The professor says that A because …. That is how the opinion of the reading is supposed heavily.
Furthermore, the reading conceives that 2. Likewise, the professor asserts that B, which sustains the reading as well. He believes that ….
Finally, the professor in the listening material maintains that C since …, just as the it is said in the reading that 3. Therefore, the reading passage has been propped up again.
Therefore, the speaker agrees totally with the opinion on the topics made in the reading.
0 应填入学术短文探讨的话题
1、2、3 简单重述学术短文的分论点（尽可能不要抄袭原文）
A、B、C 填写听力材料的分论点和论据
模糊厌恶（Ambiguity aversion），意指在所有伴随风险的不确定性中下注的话，人们倾向于下注已知的不确定类型，而不是未知类型。比如，盒子里有100个球，你愿意赌能够随机摸出的球为红色——如果已知有50个在其中的话。如果盒子里有什么颜色的球全然未知，那么你不会愿意为之下注红色的。一般的风险事情具有确定的概率计算，而暧昧事件则具有更大的不确定性，因为概率分布本身可能就是未知的。
奈特氏不确定性（英语：Knightian uncertainty），指无法被衡量期望值、不能被计算或然率、无法被预知的风险。
In game theory, cheap talk is communication between players that does not directly affect the payoffs of the game. Providing and receiving information is free. This is in contrast to signaling in which sending certain messages may be costly for the sender depending on the state of the world.
